@@ -0,0 +1,34 @@
+# Keep the build context small and fast.
+.git
+.venv
+venv
+
+# Python caches / build artifacts
+__pycache__/
+*.py[cod]
+*.egg-info/
+.eggs/
+dist/
+build/
+
+# Tooling caches
+.pytest_cache/
+.ruff_cache/
+.mypy_cache/
+.coverage
+htmlcov/
+
+# Local state and secrets (never baked into the image)
+.env
+data/*.db
+data/*.db-*
+
+# Tests are not needed at runtime
+tests/
+
+# IDE / OS
+.idea/
+.vscode/
+*.swp
+.DS_Store
+Thumbs.db

@@ -0,0 +1,36 @@
+APP_NAME=Rastro
+APP_DESCRIPTION=Observatorio publico de trackers BitTorrent
+APP_VERSION=0.1.0
+PUBLIC_URL=http://localhost:8090
+REPOSITORY_URL=
+DATABASE_URL=sqlite+aiosqlite:///data/rastro.db
+HOST=0.0.0.0
+PORT=8090
+LOG_LEVEL=INFO
+RUN_SCHEDULER=true
+# Atualizacao semanal das listas (cron em UTC). Dia: mon..sun ou 0..6.
+IMPORT_DAY_OF_WEEK=sun
+IMPORT_HOUR_UTC=4
+# Piso de garantia: se o ultimo import bem-sucedido for mais antigo que isto,
+# um import roda na inicializacao (cobre a instancia fora do ar no horario).
+IMPORT_MAX_AGE_HOURS=144
+DEFAULT_PROBE_INTERVAL_SECONDS=10800
+MIN_PROBE_INTERVAL_SECONDS=1800
+PROBE_CONCURRENCY=10
+PROBE_TIMEOUT_SECONDS=10
+MAX_RESPONSE_BYTES=1048576
+RAW_RESULT_RETENTION_DAYS=90
+ENABLE_IPV4=true
+ENABLE_IPV6=auto
+GEOIP_ASN_DB=
+
+# Fontes de lista: os padroes ficam no codigo (app/config.py). Descomente e
+# defina uma linha apenas para sobrescrever a URL correspondente.
+#TRACKERSLIST_ALL_URL=
+#TRACKERSLIST_BEST_URL=
+#NEWTRACKON_ALL_URL=
+#NEWTRACKON_STABLE_URL=
+#NEWTRACKON_LIVE_URL=
+#JOKEPOOL710_ALL_URL=
+#XIU2_ALL_URL=
+#XIU2_BEST_URL=

@@ -0,0 +1,43 @@
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.egg-info/
+.eggs/
+dist/
+build/
+.venv/
+venv/
+
+# Secrets — keep .env.example tracked
+.env
+.env.local
+.env.*.local
+
+# SQLite / local state
+data/*.db
+data/*.db-*
+*.sqlite
+*.sqlite3
+!data/.gitkeep
+backups/
+
+# IDE / OS
+.idea/
+.vscode/
+*.swp
+.DS_Store
+Thumbs.db
+desktop.ini
+
+# Test / coverage
+.coverage
+.coverage.*
+coverage.xml
+htmlcov/
+.pytest_cache/
+.ruff_cache/
+.mypy_cache/
+
+# Logs
+*.log

@@ -0,0 +1,43 @@
+# syntax=docker/dockerfile:1
+
+FROM python:3.13-slim AS builder
+
+WORKDIR /build
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    gcc \
+    && rm -rf /var/lib/apt/lists/*
+
+COPY pyproject.toml README.md LICENSE ./
+COPY app ./app
+COPY alembic ./alembic
+COPY alembic.ini ./
+
+RUN pip install --no-cache-dir --upgrade pip \
+    && pip install --no-cache-dir .
+
+
+FROM python:3.13-slim
+
+ENV PYTHONDONTWRITEBYTECODE=1 \
+    PYTHONUNBUFFERED=1 \
+    APP_HOME=/app
+
+WORKDIR /app
+
+RUN useradd --create-home --uid 10001 --shell /usr/sbin/nologin rastro \
+    && mkdir -p /app/data \
+    && chown -R rastro:rastro /app
+
+COPY --from=builder /usr/local/lib/python3.13/site-packages /usr/local/lib/python3.13/site-packages
+COPY --from=builder /usr/local/bin /usr/local/bin
+COPY --chown=rastro:rastro app ./app
+COPY --chown=rastro:rastro alembic ./alembic
+COPY --chown=rastro:rastro alembic.ini pyproject.toml LICENSE README.md ./
+
+USER rastro
+EXPOSE 8090
+
+HEALTHCHECK --interval=30s --timeout=5s --start-period=20s --retries=3 \
+  CMD python -c "import urllib.request; urllib.request.urlopen('http://127.0.0.1:8090/healthz', timeout=3)"
+
+CMD ["python", "-m", "uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "8090", "--workers", "1"]

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Pablo Murad
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

@@ -0,0 +1,7 @@
+# Rastro
+
+![Python, FastAPI, SQLite, Docker](https://skillicons.dev/icons?i=py,fastapi,sqlite,docker)
+
+Rastro is a self-hosted box that watches public BitTorrent trackers: it pulls the lists people already publish, knocks on each `udp`/`http`/`https` announce URL, writes down whether the peer actually speaks the protocol, and serves the history as a page, JSON, and a short curated text list.
+
+WE DO NOT SHARE FILES.

@@ -0,0 +1,40 @@
+[alembic]
+script_location = alembic
+prepend_sys_path = .
+version_path_separator = os
+
+sqlalchemy.url = sqlite+aiosqlite:///data/rastro.db
+
+[loggers]
+keys = root,sqlalchemy,alembic
+
+[handlers]
+keys = console
+
+[formatters]
+keys = generic
+
+[logger_root]
+level = WARN
+handlers = console
+qualname =
+
+[logger_sqlalchemy]
+level = WARN
+handlers =
+qualname = sqlalchemy.engine
+
+[logger_alembic]
+level = INFO
+handlers =
+qualname = alembic
+
+[handler_console]
+class = StreamHandler
+args = (sys.stderr,)
+level = NOTSET
+formatter = generic
+
+[formatter_generic]
+format = %(levelname)-5.5s [%(name)s] %(message)s
+datefmt = %H:%M:%S

@@ -0,0 +1,63 @@
+from __future__ import annotations
+
+import asyncio
+from logging.config import fileConfig
+
+from alembic import context
+from app.config import get_settings
+from app.db.models import Base
+from sqlalchemy import pool
+from sqlalchemy.engine import Connection
+from sqlalchemy.ext.asyncio import async_engine_from_config
+
+config = context.config
+if config.config_file_name is not None:
+    fileConfig(config.config_file_name)
+
+target_metadata = Base.metadata
+settings = get_settings()
+config.set_main_option("sqlalchemy.url", settings.database_url)
+
+
+def run_migrations_offline() -> None:
+    url = config.get_main_option("sqlalchemy.url")
+    context.configure(
+        url=url,
+        target_metadata=target_metadata,
+        literal_binds=True,
+        dialect_opts={"paramstyle": "named"},
+        render_as_batch=True,
+    )
+    with context.begin_transaction():
+        context.run_migrations()
+
+
+def do_run_migrations(connection: Connection) -> None:
+    context.configure(
+        connection=connection,
+        target_metadata=target_metadata,
+        render_as_batch=True,
+    )
+    with context.begin_transaction():
+        context.run_migrations()
+
+
+async def run_async_migrations() -> None:
+    connectable = async_engine_from_config(
+        config.get_section(config.config_ini_section, {}),
+        prefix="sqlalchemy.",
+        poolclass=pool.NullPool,
+    )
+    async with connectable.connect() as connection:
+        await connection.run_sync(do_run_migrations)
+    await connectable.dispose()
+
+
+def run_migrations_online() -> None:
+    asyncio.run(run_async_migrations())
+
+
+if context.is_offline_mode():
+    run_migrations_offline()
+else:
+    run_migrations_online()

@@ -0,0 +1,21 @@
+"""Alembic revision template."""
+from __future__ import annotations
+
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+{{ imports if imports else "" }}
+
+revision: str = {{ repr(up_revision) }}
+down_revision: Union[str, None] = {{ repr(down_revision) }}
+branch_labels: Union[str, Sequence[str], None] = {{ repr(branch_labels) }}
+depends_on: Union[str, Sequence[str], None] = {{ repr(depends_on) }}
+
+
+def upgrade() -> None:
+    ${upgrades if upgrades else "pass"}
+
+
+def downgrade() -> None:
+    ${downgrades if downgrades else "pass"}

@@ -0,0 +1,209 @@
+"""Initial schema."""
+
+from __future__ import annotations
+
+from collections.abc import Sequence
+
+import sqlalchemy as sa
+from alembic import op
+
+revision: str = "0001_initial"
+down_revision: str | None = None
+branch_labels: str | Sequence[str] | None = None
+depends_on: str | Sequence[str] | None = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        "trackers",
+        sa.Column("id", sa.Integer(), autoincrement=True, nullable=False),
+        sa.Column("canonical_url", sa.String(length=512), nullable=False),
+        sa.Column("scheme", sa.String(length=16), nullable=False),
+        sa.Column("hostname", sa.String(length=255), nullable=False),
+        sa.Column("port", sa.Integer(), nullable=False),
+        sa.Column("path", sa.String(length=512), nullable=False),
+        sa.Column("current_status", sa.String(length=32), nullable=False),
+        sa.Column(
+            "first_seen_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("(CURRENT_TIMESTAMP)"),
+            nullable=False,
+        ),
+        sa.Column(
+            "last_seen_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("(CURRENT_TIMESTAMP)"),
+            nullable=False,
+        ),
+        sa.Column("last_checked_at", sa.DateTime(timezone=True), nullable=True),
+        sa.Column("next_check_at", sa.DateTime(timezone=True), nullable=True),
+        sa.Column("announced_interval_seconds", sa.Integer(), nullable=True),
+        sa.Column("last_latency_ms", sa.Float(), nullable=True),
+        sa.Column("supports_ipv4", sa.Boolean(), nullable=True),
+        sa.Column("supports_ipv6", sa.String(length=32), nullable=True),
+        sa.Column("terminal_cname", sa.String(length=255), nullable=True),
+        sa.Column("infrastructure_fingerprint", sa.String(length=128), nullable=True),
+        sa.Column("asn", sa.Integer(), nullable=True),
+        sa.Column("network_name", sa.String(length=255), nullable=True),
+        sa.Column("country_code", sa.String(length=8), nullable=True),
+        sa.Column("consecutive_failures", sa.Integer(), nullable=False),
+        sa.Column(
+            "created_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("(CURRENT_TIMESTAMP)"),
+            nullable=False,
+        ),
+        sa.Column(
+            "updated_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("(CURRENT_TIMESTAMP)"),
+            nullable=False,
+        ),
+        sa.PrimaryKeyConstraint("id"),
+        sa.UniqueConstraint("canonical_url"),
+    )
+    op.create_index("ix_trackers_hostname", "trackers", ["hostname"])
+    op.create_index("ix_trackers_current_status", "trackers", ["current_status"])
+    op.create_index("ix_trackers_next_check_at", "trackers", ["next_check_at"])
+    op.create_index(
+        "ix_trackers_infrastructure_fingerprint", "trackers", ["infrastructure_fingerprint"]
+    )
+
+    op.create_table(
+        "tracker_sources",
+        sa.Column("id", sa.Integer(), autoincrement=True, nullable=False),
+        sa.Column("tracker_id", sa.Integer(), nullable=False),
+        sa.Column("source_name", sa.String(length=64), nullable=False),
+        sa.Column("source_url", sa.String(length=512), nullable=False),
+        sa.Column("is_best_at_source", sa.Boolean(), nullable=False),
+        sa.Column("is_stable_at_source", sa.Boolean(), nullable=False),
+        sa.Column("is_live_at_source", sa.Boolean(), nullable=False),
+        sa.Column(
+            "first_seen_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("(CURRENT_TIMESTAMP)"),
+            nullable=False,
+        ),
+        sa.Column(
+            "last_seen_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("(CURRENT_TIMESTAMP)"),
+            nullable=False,
+        ),
+        sa.ForeignKeyConstraint(["tracker_id"], ["trackers.id"], ondelete="CASCADE"),
+        sa.PrimaryKeyConstraint("id"),
+        sa.UniqueConstraint("tracker_id", "source_name", name="uq_tracker_source"),
+    )
+    op.create_index("ix_tracker_sources_tracker_id", "tracker_sources", ["tracker_id"])
+
+    op.create_table(
+        "probe_results",
+        sa.Column("id", sa.Integer(), autoincrement=True, nullable=False),
+        sa.Column("tracker_id", sa.Integer(), nullable=False),
+        sa.Column("checked_at", sa.DateTime(timezone=True), nullable=False),
+        sa.Column("status", sa.String(length=32), nullable=False),
+        sa.Column("latency_ms", sa.Float(), nullable=True),
+        sa.Column("ip_address", sa.String(length=64), nullable=True),
+        sa.Column("ip_family", sa.String(length=8), nullable=True),
+        sa.Column("response_valid", sa.Boolean(), nullable=False),
+        sa.Column("tracker_interval_seconds", sa.Integer(), nullable=True),
+        sa.Column("seeders", sa.Integer(), nullable=True),
+        sa.Column("leechers", sa.Integer(), nullable=True),
+        sa.Column("error_kind", sa.String(length=64), nullable=True),
+        sa.Column("error_detail", sa.String(length=512), nullable=True),
+        sa.ForeignKeyConstraint(["tracker_id"], ["trackers.id"], ondelete="CASCADE"),
+        sa.PrimaryKeyConstraint("id"),
+    )
+    op.create_index("ix_probe_results_tracker_id", "probe_results", ["tracker_id"])
+    op.create_index("ix_probe_results_checked_at", "probe_results", ["checked_at"])
+
+    op.create_table(
+        "dns_records",
+        sa.Column("id", sa.Integer(), autoincrement=True, nullable=False),
+        sa.Column("tracker_id", sa.Integer(), nullable=False),
+        sa.Column("record_type", sa.String(length=16), nullable=False),
+        sa.Column("value", sa.String(length=512), nullable=False),
+        sa.Column(
+            "first_seen_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("(CURRENT_TIMESTAMP)"),
+            nullable=False,
+        ),
+        sa.Column(
+            "last_seen_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("(CURRENT_TIMESTAMP)"),
+            nullable=False,
+        ),
+        sa.ForeignKeyConstraint(["tracker_id"], ["trackers.id"], ondelete="CASCADE"),
+        sa.PrimaryKeyConstraint("id"),
+        sa.UniqueConstraint("tracker_id", "record_type", "value", name="uq_dns_record"),
+    )
+    op.create_index("ix_dns_records_tracker_id", "dns_records", ["tracker_id"])
+
+    op.create_table(
+        "source_import_runs",
+        sa.Column("id", sa.Integer(), autoincrement=True, nullable=False),
+        sa.Column("source_name", sa.String(length=64), nullable=False),
+        sa.Column("started_at", sa.DateTime(timezone=True), nullable=False),
+        sa.Column("finished_at", sa.DateTime(timezone=True), nullable=True),
+        sa.Column("status", sa.String(length=32), nullable=False),
+        sa.Column("received_count", sa.Integer(), nullable=False),
+        sa.Column("valid_count", sa.Integer(), nullable=False),
+        sa.Column("new_count", sa.Integer(), nullable=False),
+        sa.Column("unsupported_count", sa.Integer(), nullable=False),
+        sa.Column("error_message", sa.String(length=512), nullable=True),
+        sa.Column("etag", sa.String(length=255), nullable=True),
+        sa.Column("last_modified", sa.String(length=255), nullable=True),
+        sa.PrimaryKeyConstraint("id"),
+    )
+    op.create_index("ix_source_import_runs_source_name", "source_import_runs", ["source_name"])
+
+    op.create_table(
+        "collector_cache",
+        sa.Column("id", sa.Integer(), autoincrement=True, nullable=False),
+        sa.Column("source_name", sa.String(length=64), nullable=False),
+        sa.Column("feed_url", sa.String(length=512), nullable=False),
+        sa.Column("etag", sa.String(length=255), nullable=True),
+        sa.Column("last_modified", sa.String(length=255), nullable=True),
+        sa.Column(
+            "updated_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("(CURRENT_TIMESTAMP)"),
+            nullable=False,
+        ),
+        sa.PrimaryKeyConstraint("id"),
+        sa.UniqueConstraint("feed_url"),
+    )
+
+    op.create_table(
+        "app_state",
+        sa.Column("key", sa.String(length=64), nullable=False),
+        sa.Column("value", sa.Text(), nullable=False),
+        sa.Column(
+            "updated_at",
+            sa.DateTime(timezone=True),
+            server_default=sa.text("(CURRENT_TIMESTAMP)"),
+            nullable=False,
+        ),
+        sa.PrimaryKeyConstraint("key"),
+    )
+
+
+def downgrade() -> None:
+    op.drop_table("app_state")
+    op.drop_table("collector_cache")
+    op.drop_index("ix_source_import_runs_source_name", table_name="source_import_runs")
+    op.drop_table("source_import_runs")
+    op.drop_index("ix_dns_records_tracker_id", table_name="dns_records")
+    op.drop_table("dns_records")
+    op.drop_index("ix_probe_results_checked_at", table_name="probe_results")
+    op.drop_index("ix_probe_results_tracker_id", table_name="probe_results")
+    op.drop_table("probe_results")
+    op.drop_index("ix_tracker_sources_tracker_id", table_name="tracker_sources")
+    op.drop_table("tracker_sources")
+    op.drop_index("ix_trackers_infrastructure_fingerprint", table_name="trackers")
+    op.drop_index("ix_trackers_next_check_at", table_name="trackers")
+    op.drop_index("ix_trackers_current_status", table_name="trackers")
+    op.drop_index("ix_trackers_hostname", table_name="trackers")
+    op.drop_table("trackers")

@@ -0,0 +1,3 @@
+__all__ = ["__version__"]
+
+__version__ = "0.1.0"

@@ -0,0 +1,4 @@
+from app.cli import main
+
+if __name__ == "__main__":
+    raise SystemExit(main())

@@ -0,0 +1,247 @@
+from __future__ import annotations
+
+from datetime import UTC, datetime
+from pathlib import Path
+from typing import Any
+
+from fastapi import APIRouter, Depends, Query, Request
+from fastapi.responses import HTMLResponse, RedirectResponse
+from fastapi.templating import Jinja2Templates
+from sqlalchemy import func, select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.config import Settings, get_settings
+from app.db import get_session
+from app.db.models import ProbeResult, Tracker
+from app.db.queries import get_app_state, get_tracker, overview_stats, trackers_query
+from app.i18n import (
+    html_lang,
+    lang_redirect,
+    render_bootstrap_note,
+    resolve_lang,
+    translate,
+    translator,
+)
+from app.services.scoring import load_metrics
+from app.services.smart_list import SmartListFilters, build_smart_list, format_smart_list_txt
+
+router = APIRouter()
+TEMPLATES_DIR = Path(__file__).resolve().parent.parent / "templates"
+templates = Jinja2Templates(directory=str(TEMPLATES_DIR))
+
+
+def _fmt_pct(value: float | None) -> str:
+    if value is None:
+        return "—"
+    return f"{value * 100:.1f}%"
+
+
+def _fmt_ms(value: float | None) -> str:
+    if value is None:
+        return "—"
+    return f"{value:.0f} ms"
+
+
+def _fmt_dt(value: datetime | None) -> str:
+    if value is None:
+        return "—"
+    if value.tzinfo is None:
+        value = value.replace(tzinfo=UTC)
+    return value.strftime("%Y-%m-%d %H:%M UTC")
+
+
+templates.env.filters["pct"] = _fmt_pct
+templates.env.filters["ms"] = _fmt_ms
+templates.env.filters["dt"] = _fmt_dt
+
+
+def _base_ctx(request: Request, settings: Settings, **extra: Any) -> dict[str, Any]:
+    lang = resolve_lang(request)
+    ctx = {
+        "request": request,
+        "app_name": settings.app_name,
+        "app_description": translate(lang, "tagline"),
+        "app_version": settings.app_version,
+        "repository_url": settings.repository_url,
+        "lang": lang,
+        "html_lang": html_lang(lang),
+        "t": translator(lang),
+    }
+    ctx.update(extra)
+    return ctx
+
+
+@router.get("/lang/{code}")
+async def set_lang(code: str, request: Request) -> RedirectResponse:
+    return lang_redirect(request, code)
+
+
+@router.get("/", response_class=HTMLResponse)
+async def home(
+    request: Request,
+    session: AsyncSession = Depends(get_session),
+    settings: Settings = Depends(get_settings),
+    q: str | None = None,
+    scheme: str | None = None,
+    status: str | None = None,
+    source: str | None = None,
+    sort: str = "checked",
+    order: str = "desc",
+    page: int = Query(1, ge=1),
+) -> HTMLResponse:
+    per_page = 50
+    stats = await overview_stats(session)
+    lang = resolve_lang(request)
+    bootstrap_note = render_bootstrap_note(lang, await get_app_state(session, "bootstrap_note"))
+    stmt = trackers_query(q=q, scheme=scheme, status=status, source=source, sort=sort, order=order)
+    total = int(
+        await session.scalar(select(func.count()).select_from(stmt.order_by(None).subquery())) or 0
+    )
+    rows = await session.execute(stmt.offset((page - 1) * per_page).limit(per_page))
+    trackers = list(rows.scalars().unique().all())
+
+    table_rows = []
+    for t in trackers:
+        metrics = await load_metrics(session, t)
+        alias_count = 0
+        if t.infrastructure_fingerprint:
+            alias_count = int(
+                await session.scalar(
+                    select(func.count())
+                    .select_from(Tracker)
+                    .where(
+                        Tracker.infrastructure_fingerprint == t.infrastructure_fingerprint,
+                        Tracker.id != t.id,
+                    )
+                )
+                or 0
+            )
+        table_rows.append(
+            {
+                "tracker": t,
+                "metrics": metrics,
+                "alias_count": alias_count,
+                "sources": ", ".join(sorted({s.source_name for s in t.sources})),
+            }
+        )
+
+    return templates.TemplateResponse(
+        request,
+        "index.html",
+        _base_ctx(
+            request,
+            settings,
+            stats=stats,
+            bootstrap_note=bootstrap_note,
+            rows=table_rows,
+            q=q or "",
+            scheme=scheme or "",
+            status=status or "",
+            source=source or "",
+            sort=sort,
+            order=order,
+            page=page,
+            total=total,
+            per_page=per_page,
+            pages=max(1, (total + per_page - 1) // per_page),
+            active_nav="trackers",
+        ),
+    )
+
+
+@router.get("/trackers/{tracker_id}", response_class=HTMLResponse)
+async def tracker_page(
+    tracker_id: int,
+    request: Request,
+    session: AsyncSession = Depends(get_session),
+    settings: Settings = Depends(get_settings),
+) -> HTMLResponse:
+    tracker = await get_tracker(session, tracker_id)
+    if tracker is None:
+        return HTMLResponse(translate(resolve_lang(request), "not_found"), status_code=404)
+    metrics = await load_metrics(session, tracker)
+    history = await session.execute(
+        select(ProbeResult)
+        .where(ProbeResult.tracker_id == tracker.id)
+        .order_by(ProbeResult.checked_at.desc())
+        .limit(40)
+    )
+    probes = list(history.scalars().all())
+    aliases: list[Tracker] = []
+    if tracker.infrastructure_fingerprint:
+        alias_rows = await session.execute(
+            select(Tracker)
+            .where(
+                Tracker.infrastructure_fingerprint == tracker.infrastructure_fingerprint,
+                Tracker.id != tracker.id,
+            )
+            .limit(20)
+        )
+        aliases = list(alias_rows.scalars().all())
+
+    # oldest→newest so the sparkline draws left to right
+    spark = [p.latency_ms or 0 for p in reversed(probes) if p.latency_ms is not None][-24:]
+
+    return templates.TemplateResponse(
+        request,
+        "tracker_detail.html",
+        _base_ctx(
+            request,
+            settings,
+            tracker=tracker,
+            metrics=metrics,
+            probes=probes,
+            aliases=aliases,
+            spark=spark,
+            active_nav="trackers",
+        ),
+    )
+
+
+@router.get("/lists", response_class=HTMLResponse)
+async def lists_page(
+    request: Request,
+    session: AsyncSession = Depends(get_session),
+    settings: Settings = Depends(get_settings),
+    protocol: str | None = None,
+    min_uptime: float = 0.95,
+    max_latency: float | None = None,
+    ip: str = "any",
+    limit: int = Query(20, ge=1, le=50),
+    min_age_days: float = 0.0,
+    diversity: bool = True,
+) -> HTMLResponse:
+    filters = SmartListFilters(
+        protocol=protocol or None,
+        min_uptime=min_uptime,
+        max_latency_ms=max_latency,
+        ip_version=ip,
+        limit=limit,
+        min_age_days=min_age_days,
+        diversity=diversity,
+    )
+    entries = await build_smart_list(session, filters)
+    txt = format_smart_list_txt(entries)
+    qs = str(request.url.query)
+    api_txt = f"/api/v1/lists/smart.txt?{qs}" if qs else "/api/v1/lists/smart.txt"
+    api_json = f"/api/v1/lists/smart.json?{qs}" if qs else "/api/v1/lists/smart.json"
+    return templates.TemplateResponse(
+        request,
+        "lists.html",
+        _base_ctx(
+            request,
+            settings,
+            entries=entries,
+            txt=txt,
+            protocol=protocol or "",
+            min_uptime=min_uptime,
+            max_latency=max_latency if max_latency is not None else "",
+            ip=ip,
+            limit=limit,
+            min_age_days=min_age_days,
+            diversity=diversity,
+            api_txt=api_txt,
+            api_json=api_json,
+            active_nav="lists",
+        ),
+    )

@@ -0,0 +1,229 @@
+from __future__ import annotations
+
+import hashlib
+from datetime import UTC, datetime
+from typing import Any
+
+from fastapi import APIRouter, Depends, HTTPException, Query, Request, Response
+from sqlalchemy import func, select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.config import Settings, get_settings
+from app.db import get_session
+from app.db.models import Tracker
+from app.db.queries import get_tracker, trackers_query
+from app.services.scoring import load_metrics
+from app.services.smart_list import (
+    SmartListFilters,
+    build_live_list,
+    build_smart_list,
+    format_smart_list_json,
+    format_smart_list_txt,
+)
+
+router = APIRouter()
+
+
+def _iso(dt: datetime | None) -> str | None:
+    if dt is None:
+        return None
+    if dt.tzinfo is None:
+        dt = dt.replace(tzinfo=UTC)
+    return dt.isoformat().replace("+00:00", "Z")
+
+
+@router.get("/meta")
+async def meta(settings: Settings = Depends(get_settings)) -> dict[str, Any]:
+    return {
+        "name": settings.app_name,
+        "description": settings.app_description,
+        "version": settings.app_version,
+        "public_url": settings.public_url,
+        "repository_url": settings.repository_url or None,
+        "timezone": "UTC",
+        "generated_at": datetime.now(UTC).isoformat().replace("+00:00", "Z"),
+    }
+
+
+@router.get("/trackers")
+async def list_trackers(
+    session: AsyncSession = Depends(get_session),
+    q: str | None = None,
+    scheme: str | None = None,
+    status: str | None = None,
+    source: str | None = None,
+    sort: str = "checked",
+    order: str = "desc",
+    page: int = Query(1, ge=1),
+    per_page: int = Query(50, ge=1, le=200),
+) -> dict[str, Any]:
+    stmt = trackers_query(q=q, scheme=scheme, status=status, source=source, sort=sort, order=order)
+    count_stmt = select(func.count()).select_from(stmt.order_by(None).subquery())
+    total = int(await session.scalar(count_stmt) or 0)
+    rows = await session.execute(stmt.offset((page - 1) * per_page).limit(per_page))
+    trackers = list(rows.scalars().unique().all())
+    items = []
+    for t in trackers:
+        metrics = await load_metrics(session, t)
+        items.append(_tracker_dict(t, metrics))
+    return {
+        "total": total,
+        "page": page,
+        "per_page": per_page,
+        "items": items,
+    }
+
+
+@router.get("/trackers/{tracker_id}")
+async def tracker_detail(
+    tracker_id: int,
+    session: AsyncSession = Depends(get_session),
+) -> dict[str, Any]:
+    tracker = await get_tracker(session, tracker_id)
+    if tracker is None:
+        raise HTTPException(status_code=404, detail="not found")
+    metrics = await load_metrics(session, tracker)
+    data = _tracker_dict(tracker, metrics)
+    data["sources"] = [
+        {
+            "name": s.source_name,
+            "is_best_at_source": s.is_best_at_source,
+            "is_stable_at_source": s.is_stable_at_source,
+            "is_live_at_source": s.is_live_at_source,
+            "last_seen_at": _iso(s.last_seen_at),
+        }
+        for s in tracker.sources
+    ]
+    data["dns_records"] = [
+        {"type": d.record_type, "value": d.value, "last_seen_at": _iso(d.last_seen_at)}
+        for d in tracker.dns_records
+    ]
+    return data
+
+
+def _tracker_dict(t: Tracker, metrics) -> dict[str, Any]:
+    return {
+        "id": t.id,
+        "url": t.canonical_url,
+        "scheme": t.scheme,
+        "hostname": t.hostname,
+        "port": t.port,
+        "status": t.current_status,
+        "score": metrics.score,
+        "provisional": metrics.provisional,
+        "uptime_24h": metrics.uptime_24h,
+        "uptime_7d": metrics.uptime_7d,
+        "uptime_30d": metrics.uptime_30d,
+        "latency_median_7d": metrics.latency_median_7d,
+        "latency_p95_7d": metrics.latency_p95_7d,
+        "valid_rate_7d": metrics.valid_rate_7d,
+        "measurement_count": metrics.measurement_count,
+        "supports_ipv4": t.supports_ipv4,
+        "supports_ipv6": t.supports_ipv6,
+        "asn": t.asn,
+        "network_name": t.network_name,
+        "country_code": t.country_code,
+        "infrastructure_fingerprint": t.infrastructure_fingerprint,
+        "terminal_cname": t.terminal_cname,
+        "last_checked_at": _iso(t.last_checked_at),
+        "first_seen_at": _iso(t.first_seen_at),
+        "sources": [s.source_name for s in (t.sources or [])],
+    }
+
+
+def _etag_for(content: str | bytes) -> str:
+    if isinstance(content, str):
+        content = content.encode("utf-8")
+    return '"' + hashlib.sha256(content).hexdigest()[:32] + '"'
+
+
+@router.get("/lists/smart.txt")
+async def smart_list_txt(
+    request: Request,
+    session: AsyncSession = Depends(get_session),
+    protocol: str | None = None,
+    min_uptime: float = 0.95,
+    max_latency: float | None = None,
+    ip: str = "any",
+    limit: int = Query(20, ge=1, le=50),
+    min_age_days: float = 0.0,
+    diversity: bool = True,
+) -> Response:
+    filters = SmartListFilters(
+        protocol=protocol,
+        min_uptime=min_uptime,
+        max_latency_ms=max_latency,
+        ip_version=ip,
+        limit=limit,
+        min_age_days=min_age_days,
+        diversity=diversity,
+    )
+    entries = await build_smart_list(session, filters)
+    body = format_smart_list_txt(entries)
+    etag = _etag_for(body)
+    if request.headers.get("if-none-match") == etag:
+        return Response(
+            status_code=304, headers={"ETag": etag, "Cache-Control": "public, max-age=60"}
+        )
+    return Response(
+        content=body,
+        media_type="text/plain; charset=utf-8",
+        headers={"ETag": etag, "Cache-Control": "public, max-age=60"},
+    )
+
+
+@router.get("/lists/smart.json")
+async def smart_list_json(
+    request: Request,
+    session: AsyncSession = Depends(get_session),
+    protocol: str | None = None,
+    min_uptime: float = 0.95,
+    max_latency: float | None = None,
+    ip: str = "any",
+    limit: int = Query(20, ge=1, le=50),
+    min_age_days: float = 0.0,
+    diversity: bool = True,
+) -> Response:
+    filters = SmartListFilters(
+        protocol=protocol,
+        min_uptime=min_uptime,
+        max_latency_ms=max_latency,
+        ip_version=ip,
+        limit=limit,
+        min_age_days=min_age_days,
+        diversity=diversity,
+    )
+    entries = await build_smart_list(session, filters)
+    payload = format_smart_list_json(entries)
+    import orjson
+
+    body = orjson.dumps(payload)
+    etag = _etag_for(body)
+    if request.headers.get("if-none-match") == etag:
+        return Response(
+            status_code=304, headers={"ETag": etag, "Cache-Control": "public, max-age=60"}
+        )
+    return Response(
+        content=body,
+        media_type="application/json",
+        headers={"ETag": etag, "Cache-Control": "public, max-age=60"},
+    )
+
+
+@router.get("/lists/live.txt")
+async def live_list_txt(
+    request: Request,
+    session: AsyncSession = Depends(get_session),
+) -> Response:
+    urls = await build_live_list(session)
+    body = ("\n\n".join(urls) + "\n") if urls else ""
+    etag = _etag_for(body)
+    if request.headers.get("if-none-match") == etag:
+        return Response(
+            status_code=304, headers={"ETag": etag, "Cache-Control": "public, max-age=30"}
+        )
+    return Response(
+        content=body,
+        media_type="text/plain; charset=utf-8",
+        headers={"ETag": etag, "Cache-Control": "public, max-age=30"},
+    )

@@ -0,0 +1,73 @@
+from __future__ import annotations
+
+import argparse
+import asyncio
+import logging
+import sys
+
+from app.collectors import import_all_sources
+from app.config import get_settings
+from app.db import get_engine, get_session_factory, init_db
+from app.db.queries import prune_probe_results
+from app.probes import probe_due_batch
+
+
+def _setup() -> None:
+    logging.basicConfig(level=logging.INFO, format="%(levelname)s %(message)s")
+    get_settings()
+    get_engine()
+
+
+async def cmd_import_sources() -> int:
+    _setup()
+    await init_db()
+    factory = get_session_factory()
+    async with factory() as session:
+        results = await import_all_sources(session)
+    for name, stats in results.items():
+        print(f"{name}: {stats}")
+    return 0
+
+
+async def cmd_probe_once() -> int:
+    _setup()
+    await init_db()
+    settings = get_settings()
+    factory = get_session_factory()
+    async with factory() as session:
+        count = await probe_due_batch(session, settings)
+    print(f"probed {count} trackers")
+    return 0
+
+
+async def cmd_prune() -> int:
+    _setup()
+    await init_db()
+    settings = get_settings()
+    factory = get_session_factory()
+    async with factory() as session:
+        removed = await prune_probe_results(session, settings.raw_result_retention_days)
+    print(f"removed {removed} probe results")
+    return 0
+
+
+def main(argv: list[str] | None = None) -> int:
+    parser = argparse.ArgumentParser(prog="python -m app.cli")
+    sub = parser.add_subparsers(dest="command", required=True)
+    sub.add_parser("import-sources", help="Import tracker sources once")
+    sub.add_parser("probe-once", help="Probe due trackers once")
+    sub.add_parser("prune", help="Prune expired raw probe results")
+    args = parser.parse_args(argv)
+
+    if args.command == "import-sources":
+        return asyncio.run(cmd_import_sources())
+    if args.command == "probe-once":
+        return asyncio.run(cmd_probe_once())
+    if args.command == "prune":
+        return asyncio.run(cmd_prune())
+    parser.error("unknown command")
+    return 2
+
+
+if __name__ == "__main__":
+    sys.exit(main())

@@ -0,0 +1,52 @@
+from __future__ import annotations
+
+import logging
+
+import httpx
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.collectors.jokepool710 import Jokepool710Collector
+from app.collectors.newtrackon import NewTrackonCollector
+from app.collectors.trackerslist import TrackerslistCollector
+from app.collectors.xiu2 import Xiu2Collector
+from app.config import Settings, get_settings
+
+logger = logging.getLogger(__name__)
+
+
+async def import_all_sources(
+    session: AsyncSession,
+    settings: Settings | None = None,
+    client: httpx.AsyncClient | None = None,
+) -> dict[str, object]:
+    settings = settings or get_settings()
+    collectors = [
+        TrackerslistCollector(),
+        NewTrackonCollector(),
+        Jokepool710Collector(),
+        Xiu2Collector(),
+    ]
+    owns_client = client is None
+    if client is None:
+        client = httpx.AsyncClient(
+            timeout=settings.probe_timeout_seconds,
+            headers={"User-Agent": settings.user_agent()},
+            follow_redirects=True,
+        )
+    results: dict[str, object] = {}
+    try:
+        for collector in collectors:
+            stats = await collector.import_once(session, client, settings)
+            results[collector.name] = stats
+            logger.info(
+                "import %s: received=%s valid=%s new=%s error=%s",
+                collector.name,
+                stats.received,
+                stats.valid,
+                stats.new,
+                stats.error,
+            )
+    finally:
+        if owns_client:
+            await client.aclose()
+    return results

@@ -0,0 +1,351 @@
+from __future__ import annotations
+
+from collections.abc import Sequence
+from dataclasses import dataclass, field
+from datetime import UTC, datetime
+from typing import Protocol
+
+import httpx
+from sqlalchemy import select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.config import Settings
+from app.db.models import CollectorCache, SourceImportRun, Tracker, TrackerSource
+from app.services.normalize import (
+    NormalizationError,
+    NormalizedTracker,
+    normalize_tracker_url,
+    parse_tracker_list_text,
+)
+
+
+@dataclass
+class ImportStats:
+    received: int = 0
+    valid: int = 0
+    new: int = 0
+    unsupported: int = 0
+    skipped_invalid: int = 0
+    not_modified: bool = False
+    error: str | None = None
+
+
+@dataclass
+class FeedFetchResult:
+    body: str | None
+    etag: str | None
+    last_modified: str | None
+    not_modified: bool = False
+    status_code: int = 0
+
+
+async def fetch_text_feed(
+    client: httpx.AsyncClient,
+    url: str,
+    *,
+    etag: str | None = None,
+    last_modified: str | None = None,
+) -> FeedFetchResult:
+    headers: dict[str, str] = {}
+    if etag:
+        headers["If-None-Match"] = etag
+    if last_modified:
+        headers["If-Modified-Since"] = last_modified
+    response = await client.get(url, headers=headers)
+    if response.status_code == 304:
+        return FeedFetchResult(
+            body=None,
+            etag=etag,
+            last_modified=last_modified,
+            not_modified=True,
+            status_code=304,
+        )
+    response.raise_for_status()
+    return FeedFetchResult(
+        body=response.text,
+        etag=response.headers.get("ETag"),
+        last_modified=response.headers.get("Last-Modified"),
+        not_modified=False,
+        status_code=response.status_code,
+    )
+
+
+async def get_cache(session: AsyncSession, feed_url: str) -> CollectorCache | None:
+    result = await session.execute(
+        select(CollectorCache).where(CollectorCache.feed_url == feed_url)
+    )
+    return result.scalar_one_or_none()
+
+
+async def upsert_cache(
+    session: AsyncSession,
+    *,
+    source_name: str,
+    feed_url: str,
+    etag: str | None,
+    last_modified: str | None,
+) -> None:
+    row = await get_cache(session, feed_url)
+    now = datetime.now(UTC)
+    if row is None:
+        session.add(
+            CollectorCache(
+                source_name=source_name,
+                feed_url=feed_url,
+                etag=etag,
+                last_modified=last_modified,
+                updated_at=now,
+            )
+        )
+    else:
+        row.etag = etag
+        row.last_modified = last_modified
+        row.updated_at = now
+
+
+def sanitize_error(message: str, limit: int = 512) -> str:
+    text = " ".join(message.split())
+    return text[:limit]
+
+
+async def upsert_tracker_from_normalized(
+    session: AsyncSession,
+    normalized: NormalizedTracker,
+    *,
+    source_name: str,
+    source_url: str,
+    is_best: bool = False,
+    is_stable: bool = False,
+    is_live: bool = False,
+) -> tuple[Tracker, bool]:
+    """Insert or update tracker + source association. Returns (tracker, created)."""
+    now = datetime.now(UTC)
+    result = await session.execute(
+        select(Tracker).where(Tracker.canonical_url == normalized.canonical_url)
+    )
+    tracker = result.scalar_one_or_none()
+    created = False
+    if tracker is None:
+        created = True
+        tracker = Tracker(
+            canonical_url=normalized.canonical_url,
+            scheme=normalized.scheme,
+            hostname=normalized.hostname,
+            port=normalized.port,
+            path=normalized.path,
+            current_status="unsupported" if normalized.unsupported else "unknown",
+            first_seen_at=now,
+            last_seen_at=now,
+            consecutive_failures=0,
+        )
+        session.add(tracker)
+        await session.flush()
+    else:
+        tracker.last_seen_at = now
+        if normalized.unsupported:
+            tracker.current_status = "unsupported"
+
+    src_result = await session.execute(
+        select(TrackerSource).where(
+            TrackerSource.tracker_id == tracker.id,
+            TrackerSource.source_name == source_name,
+        )
+    )
+    association = src_result.scalar_one_or_none()
+    if association is None:
+        session.add(
+            TrackerSource(
+                tracker_id=tracker.id,
+                source_name=source_name,
+                source_url=source_url,
+                is_best_at_source=is_best,
+                is_stable_at_source=is_stable,
+                is_live_at_source=is_live,
+                first_seen_at=now,
+                last_seen_at=now,
+            )
+        )
+    else:
+        association.last_seen_at = now
+        association.source_url = source_url
+        if is_best:
+            association.is_best_at_source = True
+        if is_stable:
+            association.is_stable_at_source = True
+        if is_live:
+            association.is_live_at_source = True
+    return tracker, created
+
+
+@dataclass
+class ParsedEntries:
+    valid: list[NormalizedTracker] = field(default_factory=list)
+    unsupported: list[NormalizedTracker] = field(default_factory=list)
+    invalid: int = 0
+    received: int = 0
+
+
+def parse_and_normalize(body: str) -> ParsedEntries:
+    raw_urls = parse_tracker_list_text(body)
+    parsed = ParsedEntries(received=len(raw_urls))
+    seen_canonical: set[str] = set()
+    for raw in raw_urls:
+        try:
+            normalized = normalize_tracker_url(raw)
+        except NormalizationError:
+            parsed.invalid += 1
+            continue
+        if normalized.canonical_url in seen_canonical:
+            continue
+        seen_canonical.add(normalized.canonical_url)
+        if normalized.unsupported:
+            parsed.unsupported.append(normalized)
+        else:
+            parsed.valid.append(normalized)
+    return parsed
+
+
+@dataclass
+class EnrichmentFeed:
+    """A secondary feed that only sets provenance flags (best/stable/live).
+
+    Enrichment feeds are best-effort: a failure never fails the whole import and
+    never touches measurement data.
+    """
+
+    url: str
+    is_best: bool = False
+    is_stable: bool = False
+    is_live: bool = False
+
+
+async def import_source_feeds(
+    session: AsyncSession,
+    client: httpx.AsyncClient,
+    *,
+    source_name: str,
+    primary_url: str,
+    enrichments: Sequence[EnrichmentFeed] = (),
+) -> ImportStats:
+    """Run the full import lifecycle for a text-feed source.
+
+    Generalizes the pattern used by the trackerslist/newTrackon collectors:
+    one primary feed (cached via ETag/Last-Modified, guarded against empty
+    responses) plus zero or more best-effort enrichment feeds.
+    """
+    stats = ImportStats()
+    run = SourceImportRun(
+        source_name=source_name,
+        started_at=datetime.now(UTC),
+        status="running",
+        received_count=0,
+        valid_count=0,
+        new_count=0,
+        unsupported_count=0,
+    )
+    session.add(run)
+    await session.flush()
+
+    try:
+        cache = await get_cache(session, primary_url)
+        feed = await fetch_text_feed(
+            client,
+            primary_url,
+            etag=cache.etag if cache else None,
+            last_modified=cache.last_modified if cache else None,
+        )
+        if feed.not_modified:
+            stats.not_modified = True
+            run.status = "not_modified"
+            run.etag = feed.etag
+            run.last_modified = feed.last_modified
+            run.finished_at = datetime.now(UTC)
+            await session.commit()
+            return stats
+
+        if not feed.body or not feed.body.strip():
+            # Empty body must not wipe existing data.
+            run.status = "failed"
+            run.error_message = sanitize_error(f"empty response from {source_name}")
+            run.finished_at = datetime.now(UTC)
+            stats.error = run.error_message
+            await session.commit()
+            return stats
+
+        parsed = parse_and_normalize(feed.body)
+        stats.received = parsed.received
+        stats.unsupported = len(parsed.unsupported)
+        stats.skipped_invalid = parsed.invalid
+
+        for item in parsed.valid:
+            _, created = await upsert_tracker_from_normalized(
+                session,
+                item,
+                source_name=source_name,
+                source_url=primary_url,
+            )
+            stats.valid += 1
+            if created:
+                stats.new += 1
+
+        for enrichment in enrichments:
+            try:
+                enrich_feed = await fetch_text_feed(client, enrichment.url)
+                if not enrich_feed.body:
+                    continue
+                for item in parse_and_normalize(enrich_feed.body).valid:
+                    await upsert_tracker_from_normalized(
+                        session,
+                        item,
+                        source_name=source_name,
+                        source_url=enrichment.url,
+                        is_best=enrichment.is_best,
+                        is_stable=enrichment.is_stable,
+                        is_live=enrichment.is_live,
+                    )
+            except Exception:  # noqa: BLE001 — enrichment is best-effort
+                continue
+
+        await upsert_cache(
+            session,
+            source_name=source_name,
+            feed_url=primary_url,
+            etag=feed.etag,
+            last_modified=feed.last_modified,
+        )
+        run.received_count = stats.received
+        run.valid_count = stats.valid
+        run.new_count = stats.new
+        run.unsupported_count = stats.unsupported
+        run.status = "ok"
+        run.etag = feed.etag
+        run.last_modified = feed.last_modified
+        run.finished_at = datetime.now(UTC)
+        await session.commit()
+        return stats
+    except Exception as exc:  # noqa: BLE001
+        await session.rollback()
+        session.add(
+            SourceImportRun(
+                source_name=source_name,
+                started_at=datetime.now(UTC),
+                finished_at=datetime.now(UTC),
+                status="failed",
+                received_count=0,
+                valid_count=0,
+                new_count=0,
+                unsupported_count=0,
+                error_message=sanitize_error(str(exc)),
+            )
+        )
+        await session.commit()
+        stats.error = sanitize_error(str(exc))
+        return stats
+
+
+class Collector(Protocol):
+    name: str
+
+    async def import_once(
+        self, session: AsyncSession, client: httpx.AsyncClient, settings: Settings
+    ) -> ImportStats: ...

@@ -0,0 +1,30 @@
+"""jokepool710/Torrent-_Trackers collector.
+
+Independent daily public-tracker list (MIT, (c) J0KEP00L). Only the aggregate
+`trackers_all.txt` feed is consumed; the per-protocol files it also publishes
+are subsets of that aggregate.
+"""
+
+from __future__ import annotations
+
+import httpx
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.collectors.base import ImportStats, import_source_feeds
+from app.config import Settings
+
+SOURCE_NAME = "jokepool710"
+
+
+class Jokepool710Collector:
+    name = SOURCE_NAME
+
+    async def import_once(
+        self, session: AsyncSession, client: httpx.AsyncClient, settings: Settings
+    ) -> ImportStats:
+        return await import_source_feeds(
+            session,
+            client,
+            source_name=SOURCE_NAME,
+            primary_url=settings.jokepool710_all_url,
+        )

@@ -0,0 +1,27 @@
+from __future__ import annotations
+
+import httpx
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.collectors.base import EnrichmentFeed, ImportStats, import_source_feeds
+from app.config import Settings
+
+SOURCE_NAME = "newtrackon"
+
+
+class NewTrackonCollector:
+    name = SOURCE_NAME
+
+    async def import_once(
+        self, session: AsyncSession, client: httpx.AsyncClient, settings: Settings
+    ) -> ImportStats:
+        return await import_source_feeds(
+            session,
+            client,
+            source_name=SOURCE_NAME,
+            primary_url=settings.newtrackon_all_url,
+            enrichments=[
+                EnrichmentFeed(url=settings.newtrackon_stable_url, is_stable=True),
+                EnrichmentFeed(url=settings.newtrackon_live_url, is_live=True),
+            ],
+        )

@@ -0,0 +1,24 @@
+from __future__ import annotations
+
+import httpx
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.collectors.base import EnrichmentFeed, ImportStats, import_source_feeds
+from app.config import Settings
+
+SOURCE_NAME = "trackerslist"
+
+
+class TrackerslistCollector:
+    name = SOURCE_NAME
+
+    async def import_once(
+        self, session: AsyncSession, client: httpx.AsyncClient, settings: Settings
+    ) -> ImportStats:
+        return await import_source_feeds(
+            session,
+            client,
+            source_name=SOURCE_NAME,
+            primary_url=settings.trackerslist_all_url,
+            enrichments=[EnrichmentFeed(url=settings.trackerslist_best_url, is_best=True)],
+        )

@@ -0,0 +1,31 @@
+"""XIU2/TrackersListCollection collector.
+
+Daily aggregator of several upstream tracker lists (GPL-3.0). As with the
+GPLv2 trackerslist feed, only published data is consumed -- no GPL code is
+used. `all.txt` seeds endpoints; `best.txt` only sets the best-at-source flag.
+"""
+
+from __future__ import annotations
+
+import httpx
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.collectors.base import EnrichmentFeed, ImportStats, import_source_feeds
+from app.config import Settings
+
+SOURCE_NAME = "xiu2"
+
+
+class Xiu2Collector:
+    name = SOURCE_NAME
+
+    async def import_once(
+        self, session: AsyncSession, client: httpx.AsyncClient, settings: Settings
+    ) -> ImportStats:
+        return await import_source_feeds(
+            session,
+            client,
+            source_name=SOURCE_NAME,
+            primary_url=settings.xiu2_all_url,
+            enrichments=[EnrichmentFeed(url=settings.xiu2_best_url, is_best=True)],
+        )

@@ -0,0 +1,136 @@
+from __future__ import annotations
+
+from functools import lru_cache
+from typing import Literal
+
+from pydantic import Field, field_validator, model_validator
+from pydantic_settings import BaseSettings, SettingsConfigDict
+
+
+class Settings(BaseSettings):
+    model_config = SettingsConfigDict(
+        env_file=".env",
+        env_file_encoding="utf-8",
+        extra="ignore",
+        populate_by_name=True,
+    )
+
+    app_name: str = Field(default="Rastro", alias="APP_NAME")
+    app_description: str = Field(
+        default="Observatorio publico de trackers BitTorrent",
+        alias="APP_DESCRIPTION",
+    )
+    app_version: str = Field(default="0.1.0", alias="APP_VERSION")
+    public_url: str = Field(default="http://localhost:8090", alias="PUBLIC_URL")
+    repository_url: str = Field(default="", alias="REPOSITORY_URL")
+
+    database_url: str = Field(
+        default="sqlite+aiosqlite:///data/rastro.db",
+        alias="DATABASE_URL",
+    )
+    host: str = Field(default="0.0.0.0", alias="HOST")
+    port: int = Field(default=8090, alias="PORT", ge=1, le=65535)
+    log_level: str = Field(default="INFO", alias="LOG_LEVEL")
+
+    run_scheduler: bool = Field(default=True, alias="RUN_SCHEDULER")
+    # Source lists refresh on a weekly cron (UTC). A startup import still runs
+    # when the inventory is empty or the last successful import is stale.
+    import_day_of_week: str = Field(default="sun", alias="IMPORT_DAY_OF_WEEK")
+    import_hour_utc: int = Field(default=4, alias="IMPORT_HOUR_UTC", ge=0, le=23)
+    import_max_age_hours: int = Field(default=144, alias="IMPORT_MAX_AGE_HOURS", ge=1)
+    default_probe_interval_seconds: int = Field(
+        default=10800, alias="DEFAULT_PROBE_INTERVAL_SECONDS", ge=60
+    )
+    min_probe_interval_seconds: int = Field(default=1800, alias="MIN_PROBE_INTERVAL_SECONDS", ge=60)
+    probe_concurrency: int = Field(default=10, alias="PROBE_CONCURRENCY", ge=1, le=100)
+    probe_timeout_seconds: float = Field(default=10.0, alias="PROBE_TIMEOUT_SECONDS", gt=0, le=120)
+    max_response_bytes: int = Field(default=1_048_576, alias="MAX_RESPONSE_BYTES", ge=1024)
+    raw_result_retention_days: int = Field(default=90, alias="RAW_RESULT_RETENTION_DAYS", ge=1)
+
+    enable_ipv4: bool = Field(default=True, alias="ENABLE_IPV4")
+    enable_ipv6: Literal["true", "false", "auto"] | bool = Field(
+        default="auto", alias="ENABLE_IPV6"
+    )
+    geoip_asn_db: str = Field(default="", alias="GEOIP_ASN_DB")
+
+    # Source URLs (overridable for tests)
+    trackerslist_all_url: str = Field(
+        default=("https://raw.githubusercontent.com/ngosang/trackerslist/master/trackers_all.txt"),
+        alias="TRACKERSLIST_ALL_URL",
+    )
+    trackerslist_best_url: str = Field(
+        default=("https://raw.githubusercontent.com/ngosang/trackerslist/master/trackers_best.txt"),
+        alias="TRACKERSLIST_BEST_URL",
+    )
+    newtrackon_all_url: str = Field(
+        default="https://newtrackon.com/api/all",
+        alias="NEWTRACKON_ALL_URL",
+    )
+    newtrackon_stable_url: str = Field(
+        default="https://newtrackon.com/api/stable",
+        alias="NEWTRACKON_STABLE_URL",
+    )
+    newtrackon_live_url: str = Field(
+        default="https://newtrackon.com/api/live",
+        alias="NEWTRACKON_LIVE_URL",
+    )
+    jokepool710_all_url: str = Field(
+        default=(
+            "https://raw.githubusercontent.com/jokepool710/Torrent-_Trackers/main/trackers_all.txt"
+        ),
+        alias="JOKEPOOL710_ALL_URL",
+    )
+    xiu2_all_url: str = Field(
+        default="https://raw.githubusercontent.com/XIU2/TrackersListCollection/master/all.txt",
+        alias="XIU2_ALL_URL",
+    )
+    xiu2_best_url: str = Field(
+        default="https://raw.githubusercontent.com/XIU2/TrackersListCollection/master/best.txt",
+        alias="XIU2_BEST_URL",
+    )
+
+    @field_validator("enable_ipv6", mode="before")
+    @classmethod
+    def normalize_ipv6(cls, value: object) -> object:
+        if isinstance(value, str):
+            lowered = value.strip().lower()
+            if lowered in {"1", "yes", "true"}:
+                return True
+            if lowered in {"0", "no", "false"}:
+                return False
+            if lowered == "auto":
+                return "auto"
+        return value
+
+    @model_validator(mode="after")
+    def validate_intervals(self) -> Settings:
+        if self.min_probe_interval_seconds > self.default_probe_interval_seconds:
+            raise ValueError(
+                "MIN_PROBE_INTERVAL_SECONDS cannot exceed DEFAULT_PROBE_INTERVAL_SECONDS"
+            )
+        return self
+
+    def user_agent(self) -> str:
+        base = f"{self.app_name}/{self.app_version}"
+        url = (self.public_url or "").strip()
+        if url:
+            return f"{base} (+{url})"
+        return base
+
+    def peer_id_prefix(self) -> bytes:
+        """Return a 8-byte BitTorrent peer_id style prefix derived from app name."""
+        slug = "".join(c for c in self.app_name.upper() if c.isalnum())[:6] or "RASTRO"
+        slug = slug.ljust(6, "X")[:6]
+        return f"-{slug}-".encode("ascii")
+
+    def ipv6_mode(self) -> Literal["true", "false", "auto"]:
+        if self.enable_ipv6 is True:
+            return "true"
+        if self.enable_ipv6 is False:
+            return "false"
+        return "auto"
+
+
+@lru_cache
+def get_settings() -> Settings:
+    return Settings()

@@ -0,0 +1,80 @@
+from __future__ import annotations
+
+from collections.abc import AsyncGenerator
+from pathlib import Path
+
+from sqlalchemy import event, text
+from sqlalchemy.ext.asyncio import (
+    AsyncEngine,
+    AsyncSession,
+    async_sessionmaker,
+    create_async_engine,
+)
+
+from app.config import Settings, get_settings
+
+_engine: AsyncEngine | None = None
+_session_factory: async_sessionmaker[AsyncSession] | None = None
+
+
+def _ensure_sqlite_parent(database_url: str) -> None:
+    if "sqlite" not in database_url:
+        return
+    # sqlite+aiosqlite:///data/rastro.db or ////absolute
+    raw = database_url.split("///", 1)[-1]
+    path = Path(raw)
+    if path.parent and str(path.parent) not in {".", ""}:
+        path.parent.mkdir(parents=True, exist_ok=True)
+
+
+def get_engine(settings: Settings | None = None) -> AsyncEngine:
+    global _engine, _session_factory
+    if _engine is not None:
+        return _engine
+    settings = settings or get_settings()
+    _ensure_sqlite_parent(settings.database_url)
+    _engine = create_async_engine(
+        settings.database_url,
+        echo=False,
+        connect_args={"check_same_thread": False},
+    )
+
+    @event.listens_for(_engine.sync_engine, "connect")
+    def _set_sqlite_pragma(dbapi_connection, connection_record) -> None:  # noqa: ARG001
+        cursor = dbapi_connection.cursor()
+        cursor.execute("PRAGMA journal_mode=WAL")
+        cursor.execute("PRAGMA foreign_keys=ON")
+        cursor.close()
+
+    _session_factory = async_sessionmaker(_engine, expire_on_commit=False)
+    return _engine
+
+
+def get_session_factory() -> async_sessionmaker[AsyncSession]:
+    if _session_factory is None:
+        get_engine()
+    assert _session_factory is not None
+    return _session_factory
+
+
+async def get_session() -> AsyncGenerator[AsyncSession]:
+    factory = get_session_factory()
+    async with factory() as session:
+        yield session
+
+
+async def init_db(settings: Settings | None = None) -> None:
+    """Create tables if needed (Alembic preferred; used as fallback/tests)."""
+    from app.db.models import Base
+
+    engine = get_engine(settings)
+    async with engine.begin() as conn:
+        await conn.run_sync(Base.metadata.create_all)
+        await conn.execute(text("PRAGMA journal_mode=WAL"))
+
+
+def reset_engine() -> None:
+    """Reset global engine (tests)."""
+    global _engine, _session_factory
+    _engine = None
+    _session_factory = None

@@ -0,0 +1,185 @@
+from __future__ import annotations
+
+from datetime import datetime
+
+from sqlalchemy import (
+    Boolean,
+    DateTime,
+    Float,
+    ForeignKey,
+    Index,
+    Integer,
+    String,
+    Text,
+    UniqueConstraint,
+    func,
+)
+from sqlalchemy.orm import DeclarativeBase, Mapped, mapped_column, relationship
+
+
+class Base(DeclarativeBase):
+    pass
+
+
+class Tracker(Base):
+    __tablename__ = "trackers"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    canonical_url: Mapped[str] = mapped_column(String(512), unique=True, nullable=False)
+    scheme: Mapped[str] = mapped_column(String(16), nullable=False)
+    hostname: Mapped[str] = mapped_column(String(255), nullable=False, index=True)
+    port: Mapped[int] = mapped_column(Integer, nullable=False)
+    path: Mapped[str] = mapped_column(String(512), nullable=False, default="/announce")
+
+    current_status: Mapped[str] = mapped_column(
+        String(32), nullable=False, default="unknown", index=True
+    )
+    first_seen_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), nullable=False
+    )
+    last_seen_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), nullable=False
+    )
+    last_checked_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    next_check_at: Mapped[datetime | None] = mapped_column(
+        DateTime(timezone=True), nullable=True, index=True
+    )
+    announced_interval_seconds: Mapped[int | None] = mapped_column(Integer, nullable=True)
+    last_latency_ms: Mapped[float | None] = mapped_column(Float, nullable=True)
+
+    supports_ipv4: Mapped[bool | None] = mapped_column(Boolean, nullable=True)
+    supports_ipv6: Mapped[str | None] = mapped_column(String(32), nullable=True)
+    # supports_ipv6 stores: true / false / not_tested / unknown as string for flexibility
+
+    terminal_cname: Mapped[str | None] = mapped_column(String(255), nullable=True)
+    infrastructure_fingerprint: Mapped[str | None] = mapped_column(
+        String(128), nullable=True, index=True
+    )
+    asn: Mapped[int | None] = mapped_column(Integer, nullable=True)
+    network_name: Mapped[str | None] = mapped_column(String(255), nullable=True)
+    country_code: Mapped[str | None] = mapped_column(String(8), nullable=True)
+
+    consecutive_failures: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
+    created_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), nullable=False
+    )
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True),
+        server_default=func.now(),
+        onupdate=func.now(),
+        nullable=False,
+    )
+
+    sources: Mapped[list[TrackerSource]] = relationship(back_populates="tracker")
+    probe_results: Mapped[list[ProbeResult]] = relationship(back_populates="tracker")
+    dns_records: Mapped[list[DnsRecord]] = relationship(back_populates="tracker")
+
+
+class TrackerSource(Base):
+    __tablename__ = "tracker_sources"
+    __table_args__ = (UniqueConstraint("tracker_id", "source_name", name="uq_tracker_source"),)
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    tracker_id: Mapped[int] = mapped_column(
+        ForeignKey("trackers.id", ondelete="CASCADE"), nullable=False, index=True
+    )
+    source_name: Mapped[str] = mapped_column(String(64), nullable=False)
+    source_url: Mapped[str] = mapped_column(String(512), nullable=False)
+    is_best_at_source: Mapped[bool] = mapped_column(Boolean, nullable=False, default=False)
+    is_stable_at_source: Mapped[bool] = mapped_column(Boolean, nullable=False, default=False)
+    is_live_at_source: Mapped[bool] = mapped_column(Boolean, nullable=False, default=False)
+    first_seen_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), nullable=False
+    )
+    last_seen_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), nullable=False
+    )
+
+    tracker: Mapped[Tracker] = relationship(back_populates="sources")
+
+
+class ProbeResult(Base):
+    __tablename__ = "probe_results"
+    __table_args__ = (Index("ix_probe_results_checked_at", "checked_at"),)
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    tracker_id: Mapped[int] = mapped_column(
+        ForeignKey("trackers.id", ondelete="CASCADE"), nullable=False, index=True
+    )
+    checked_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), nullable=False)
+    status: Mapped[str] = mapped_column(String(32), nullable=False)
+    latency_ms: Mapped[float | None] = mapped_column(Float, nullable=True)
+    ip_address: Mapped[str | None] = mapped_column(String(64), nullable=True)
+    ip_family: Mapped[str | None] = mapped_column(String(8), nullable=True)
+    response_valid: Mapped[bool] = mapped_column(Boolean, nullable=False, default=False)
+    tracker_interval_seconds: Mapped[int | None] = mapped_column(Integer, nullable=True)
+    seeders: Mapped[int | None] = mapped_column(Integer, nullable=True)
+    leechers: Mapped[int | None] = mapped_column(Integer, nullable=True)
+    error_kind: Mapped[str | None] = mapped_column(String(64), nullable=True)
+    error_detail: Mapped[str | None] = mapped_column(String(512), nullable=True)
+
+    tracker: Mapped[Tracker] = relationship(back_populates="probe_results")
+
+
+class DnsRecord(Base):
+    __tablename__ = "dns_records"
+    __table_args__ = (UniqueConstraint("tracker_id", "record_type", "value", name="uq_dns_record"),)
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    tracker_id: Mapped[int] = mapped_column(
+        ForeignKey("trackers.id", ondelete="CASCADE"), nullable=False, index=True
+    )
+    record_type: Mapped[str] = mapped_column(String(16), nullable=False)
+    value: Mapped[str] = mapped_column(String(512), nullable=False)
+    first_seen_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), nullable=False
+    )
+    last_seen_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), nullable=False
+    )
+
+    tracker: Mapped[Tracker] = relationship(back_populates="dns_records")
+
+
+class SourceImportRun(Base):
+    __tablename__ = "source_import_runs"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    source_name: Mapped[str] = mapped_column(String(64), nullable=False, index=True)
+    started_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), nullable=False)
+    finished_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    status: Mapped[str] = mapped_column(String(32), nullable=False, default="running")
+    received_count: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
+    valid_count: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
+    new_count: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
+    unsupported_count: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
+    error_message: Mapped[str | None] = mapped_column(String(512), nullable=True)
+    etag: Mapped[str | None] = mapped_column(String(255), nullable=True)
+    last_modified: Mapped[str | None] = mapped_column(String(255), nullable=True)
+
+
+class CollectorCache(Base):
+    """HTTP cache state per source feed URL."""
+
+    __tablename__ = "collector_cache"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    source_name: Mapped[str] = mapped_column(String(64), nullable=False)
+    feed_url: Mapped[str] = mapped_column(String(512), nullable=False, unique=True)
+    etag: Mapped[str | None] = mapped_column(String(255), nullable=True)
+    last_modified: Mapped[str | None] = mapped_column(String(255), nullable=True)
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), nullable=False
+    )
+
+
+class AppState(Base):
+    """Key/value operational state (bootstrap notes, etc.)."""
+
+    __tablename__ = "app_state"
+
+    key: Mapped[str] = mapped_column(String(64), primary_key=True)
+    value: Mapped[str] = mapped_column(Text, nullable=False)
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), nullable=False
+    )

@@ -0,0 +1,151 @@
+from __future__ import annotations
+
+from datetime import UTC, datetime, timedelta
+from typing import Any
+
+from sqlalchemy import Select, func, select
+from sqlalchemy.ext.asyncio import AsyncSession
+from sqlalchemy.orm import selectinload
+
+from app.db.models import (
+    AppState,
+    ProbeResult,
+    SourceImportRun,
+    Tracker,
+    TrackerSource,
+)
+
+
+async def get_app_state(session: AsyncSession, key: str) -> str | None:
+    row = await session.get(AppState, key)
+    return row.value if row else None
+
+
+async def set_app_state(session: AsyncSession, key: str, value: str) -> None:
+    row = await session.get(AppState, key)
+    if row is None:
+        session.add(AppState(key=key, value=value, updated_at=datetime.now(UTC)))
+    else:
+        row.value = value
+        row.updated_at = datetime.now(UTC)
+    await session.commit()
+
+
+async def count_trackers(session: AsyncSession) -> int:
+    result = await session.scalar(select(func.count()).select_from(Tracker))
+    return int(result or 0)
+
+
+async def latest_successful_import_at(session: AsyncSession) -> datetime | None:
+    """Finished-at of the most recent import that actually applied data.
+
+    `not_modified` counts as success: the upstream feed was reachable and simply
+    unchanged, so the inventory is not stale.
+    """
+    return await session.scalar(
+        select(func.max(SourceImportRun.finished_at)).where(
+            SourceImportRun.status.in_(("ok", "not_modified"))
+        )
+    )
+
+
+async def overview_stats(session: AsyncSession) -> dict[str, Any]:
+    total = await count_trackers(session)
+    by_status: dict[str, int] = {}
+    rows = await session.execute(
+        select(Tracker.current_status, func.count()).group_by(Tracker.current_status)
+    )
+    for status, count in rows.all():
+        by_status[status] = count
+
+    last_import = await session.scalar(
+        select(SourceImportRun)
+        .where(SourceImportRun.status.in_(("ok", "not_modified", "failed")))
+        .order_by(SourceImportRun.finished_at.desc())
+        .limit(1)
+    )
+    last_probe = await session.scalar(
+        select(ProbeResult).order_by(ProbeResult.checked_at.desc()).limit(1)
+    )
+
+    return {
+        "total": total,
+        "up": by_status.get("up", 0),
+        "degraded": by_status.get("degraded", 0),
+        "down": by_status.get("down", 0),
+        "unknown": by_status.get("unknown", 0),
+        "last_import_at": last_import.finished_at if last_import else None,
+        "last_probe_at": last_probe.checked_at if last_probe else None,
+    }
+
+
+def trackers_query(
+    *,
+    q: str | None = None,
+    scheme: str | None = None,
+    status: str | None = None,
+    source: str | None = None,
+    sort: str = "score",
+    order: str = "desc",
+) -> Select[tuple[Tracker]]:
+    stmt = select(Tracker).options(selectinload(Tracker.sources))
+    if q:
+        like = f"%{q.strip()}%"
+        stmt = stmt.where(Tracker.canonical_url.ilike(like) | Tracker.hostname.ilike(like))
+    if scheme:
+        stmt = stmt.where(Tracker.scheme == scheme.lower())
+    if status:
+        stmt = stmt.where(Tracker.current_status == status.lower())
+    if source:
+        stmt = stmt.join(TrackerSource).where(TrackerSource.source_name == source)
+
+    sort_map = {
+        "url": Tracker.canonical_url,
+        "status": Tracker.current_status,
+        "latency": Tracker.last_latency_ms,
+        "checked": Tracker.last_checked_at,
+        "hostname": Tracker.hostname,
+        "protocol": Tracker.scheme,
+    }
+    col = sort_map.get(sort, Tracker.last_checked_at)
+    if order.lower() == "asc":
+        stmt = stmt.order_by(col.asc().nullslast(), Tracker.id.asc())
+    else:
+        stmt = stmt.order_by(col.desc().nullslast(), Tracker.id.desc())
+    return stmt
+
+
+async def get_tracker(session: AsyncSession, tracker_id: int) -> Tracker | None:
+    result = await session.execute(
+        select(Tracker)
+        .options(
+            selectinload(Tracker.sources),
+            selectinload(Tracker.dns_records),
+        )
+        .where(Tracker.id == tracker_id)
+    )
+    return result.scalar_one_or_none()
+
+
+async def due_trackers(session: AsyncSession, limit: int = 50) -> list[Tracker]:
+    now = datetime.now(UTC)
+    result = await session.execute(
+        select(Tracker)
+        .where(
+            Tracker.current_status != "unsupported",
+            (Tracker.next_check_at.is_(None)) | (Tracker.next_check_at <= now),
+        )
+        .order_by(Tracker.next_check_at.asc().nullsfirst())
+        .limit(limit)
+    )
+    return list(result.scalars().all())
+
+
+async def prune_probe_results(session: AsyncSession, retention_days: int) -> int:
+    cutoff = datetime.now(UTC) - timedelta(days=retention_days)
+    rows = await session.execute(select(ProbeResult).where(ProbeResult.checked_at < cutoff))
+    old = list(rows.scalars().all())
+    for row in old:
+        await session.delete(row)
+    await session.commit()
+    return len(old)

@@ -0,0 +1,353 @@
+from __future__ import annotations
+
+from urllib.parse import urlparse
+
+from starlette.requests import Request
+from starlette.responses import RedirectResponse
+
+LOCALES = ("pt", "en")
+DEFAULT = "pt"
+COOKIE = "lang"
+COOKIE_MAX_AGE = 60 * 60 * 24 * 365
+
+_PT = {
+    "tagline": "Observatório público de trackers BitTorrent",
+    "nav_trackers": "Trackers",
+    "nav_trackers_sub": "lista",
+    "nav_lists": "Criar lista",
+    "nav_lists_sub": "smart",
+    "nav_github": "GitHub",
+    "nav_github_sub": "código",
+    "footer_utc": "medições em UTC",
+    "footer_disclaimer": "Não é um tracker BitTorrent · não indexa torrents",
+    "stats_label": "Resumo",
+    "stats_monitored": "monitorados",
+    "stats_up": "ativos",
+    "stats_degraded": "degradados",
+    "stats_down": "indisponíveis",
+    "stats_unknown": "desconhecidos",
+    "stats_last_import": "última importação",
+    "stats_last_probe": "última medição",
+    "filter_search": "Busca",
+    "filter_search_ph": "hostname ou URL",
+    "filter_protocol": "Protocolo",
+    "filter_any": "qualquer",
+    "filter_status": "Status",
+    "filter_source": "Origem",
+    "filter_sort": "Ordenar",
+    "sort_checked": "última verificação",
+    "sort_latency": "latência",
+    "sort_url": "URL",
+    "sort_status": "status",
+    "sort_protocol": "protocolo",
+    "filter_order": "Ordem",
+    "order_desc": "desc",
+    "order_asc": "asc",
+    "filter_submit": "Filtrar",
+    "col_tracker": "Tracker",
+    "col_protocol": "Protocolo",
+    "col_status": "Status",
+    "col_uptime7": "Uptime 7d",
+    "col_latency": "Latência med.",
+    "col_ip": "IPv4 / IPv6",
+    "col_asn": "ASN / rede",
+    "col_sources": "Fontes",
+    "col_aliases": "Aliases",
+    "col_last_check": "Última verificação",
+    "copy": "copiar",
+    "copy_btn": "Copiar",
+    "copied": "copiado",
+    "th_latency": "Latência",
+    "provisional": "provisório",
+    "empty_inventory": "Nenhum tracker no inventário ainda.",
+    "pagination": "Paginação",
+    "prev": "Anterior",
+    "next": "Próxima",
+    "page_of": "Página {page} / {pages} ({total})",
+    "lists_title": "Criar lista",
+    "lists_heading": "Smart List",
+    "lists_blurb": (
+        "Seleção de trackers estáveis e com diversidade de infraestrutura. "
+        "Classificações das fontes externas não substituem as medições próprias."
+    ),
+    "min_uptime": "Uptime mínimo (0–1)",
+    "max_latency": "Latência máxima (ms)",
+    "filter_ip": "IP",
+    "filter_limit": "Quantidade (1–50)",
+    "min_age": "Idade mínima (dias)",
+    "diversity": "Diversidade de infraestrutura",
+    "diversity_on": "ligada",
+    "diversity_off": "desligada",
+    "generate": "Gerar",
+    "selected_count": "trackers selecionados.",
+    "download_txt": "Baixar TXT",
+    "open_json": "Abrir JSON",
+    "permalink": "URL permanente",
+    "back": "voltar",
+    "score": "pontuação",
+    "score_insufficient": "insuficiente",
+    "uptime_24h": "Uptime 24h",
+    "uptime_7d": "Uptime 7d",
+    "uptime_30d": "Uptime 30d",
+    "latency_med_7d": "Latência med. 7d",
+    "latency_p95": "Latência p95 7d",
+    "valid_rate": "Respostas válidas 7d",
+    "measurements": "Medições",
+    "tracking": "Acompanhamento",
+    "days": "dias",
+    "spark_label": "Latências recentes",
+    "sources": "Origens",
+    "best_at_source": "best na origem",
+    "stable_at_source": "stable na origem",
+    "live_at_source": "live na origem",
+    "seen": "visto",
+    "no_source": "Sem associação de origem",
+    "dns": "DNS",
+    "terminal_cname": "CNAME terminal",
+    "infra_fp": "Fingerprint de infraestrutura",
+    "asn": "ASN",
+    "unknown": "desconhecido",
+    "country": "País",
+    "no_dns": "Sem registros DNS armazenados",
+    "shared_infra": "Infraestrutura compartilhada",
+    "shared_infra_blurb": (
+        "Outros endpoints com o mesmo fingerprint (sinal, não prova de operador):"
+    ),
+    "recent_history": "Histórico recente",
+    "when": "Quando (UTC)",
+    "valid": "Válida",
+    "error": "Erro",
+    "yes": "sim",
+    "no": "não",
+    "no_probes": "Sem medições ainda",
+    "not_found": "Tracker não encontrado",
+    "bootstrap_empty": (
+        "Inventário vazio: não foi possível importar fontes na inicialização ({exc}). "
+        "O site está no ar; tente novamente com `python -m app.cli import-sources`."
+    ),
+    "bootstrap_stale": (
+        "Listas desatualizadas: não foi possível importar fontes na inicialização ({exc}). "
+        "O site está no ar; tente novamente com `python -m app.cli import-sources`."
+    ),
+}
+
+_EN = {
+    "tagline": "Public observatory of BitTorrent trackers",
+    "nav_trackers": "Trackers",
+    "nav_trackers_sub": "list",
+    "nav_lists": "Build a list",
+    "nav_lists_sub": "smart",
+    "nav_github": "GitHub",
+    "nav_github_sub": "code",
+    "footer_utc": "measurements in UTC",
+    "footer_disclaimer": "Not a BitTorrent tracker · does not index torrents",
+    "stats_label": "Overview",
+    "stats_monitored": "watched",
+    "stats_up": "up",
+    "stats_degraded": "degraded",
+    "stats_down": "down",
+    "stats_unknown": "unknown",
+    "stats_last_import": "last import",
+    "stats_last_probe": "last probe",
+    "filter_search": "Search",
+    "filter_search_ph": "hostname or URL",
+    "filter_protocol": "Protocol",
+    "filter_any": "any",
+    "filter_status": "Status",
+    "filter_source": "Source",
+    "filter_sort": "Sort",
+    "sort_checked": "last check",
+    "sort_latency": "latency",
+    "sort_url": "URL",
+    "sort_status": "status",
+    "sort_protocol": "protocol",
+    "filter_order": "Order",
+    "order_desc": "desc",
+    "order_asc": "asc",
+    "filter_submit": "Filter",
+    "col_tracker": "Tracker",
+    "col_protocol": "Protocol",
+    "col_status": "Status",
+    "col_uptime7": "Uptime 7d",
+    "col_latency": "Median latency",
+    "col_ip": "IPv4 / IPv6",
+    "col_asn": "ASN / net",
+    "col_sources": "Sources",
+    "col_aliases": "Aliases",
+    "col_last_check": "Last check",
+    "copy": "copy",
+    "copy_btn": "Copy",
+    "copied": "copied",
+    "th_latency": "Latency",
+    "provisional": "provisional",
+    "empty_inventory": "No trackers in the inventory yet.",
+    "pagination": "Pagination",
+    "prev": "Previous",
+    "next": "Next",
+    "page_of": "Page {page} / {pages} ({total})",
+    "lists_title": "Build a list",
+    "lists_heading": "Smart List",
+    "lists_blurb": (
+        "Stable trackers with some infrastructure spread. "
+        "Labels from the upstream lists do not replace our own probes."
+    ),
+    "min_uptime": "Min uptime (0–1)",
+    "max_latency": "Max latency (ms)",
+    "filter_ip": "IP",
+    "filter_limit": "Count (1–50)",
+    "min_age": "Min age (days)",
+    "diversity": "Infrastructure diversity",
+    "diversity_on": "on",
+    "diversity_off": "off",
+    "generate": "Build",
+    "selected_count": "trackers selected.",
+    "download_txt": "Download TXT",
+    "open_json": "Open JSON",
+    "permalink": "Permalink",
+    "back": "back",
+    "score": "score",
+    "score_insufficient": "not enough data",
+    "uptime_24h": "Uptime 24h",
+    "uptime_7d": "Uptime 7d",
+    "uptime_30d": "Uptime 30d",
+    "latency_med_7d": "Median latency 7d",
+    "latency_p95": "p95 latency 7d",
+    "valid_rate": "Valid replies 7d",
+    "measurements": "Probes",
+    "tracking": "Watched for",
+    "days": "days",
+    "spark_label": "Recent latencies",
+    "sources": "Sources",
+    "best_at_source": "best at source",
+    "stable_at_source": "stable at source",
+    "live_at_source": "live at source",
+    "seen": "seen",
+    "no_source": "No source association",
+    "dns": "DNS",
+    "terminal_cname": "Terminal CNAME",
+    "infra_fp": "Infrastructure fingerprint",
+    "asn": "ASN",
+    "unknown": "unknown",
+    "country": "Country",
+    "no_dns": "No stored DNS records",
+    "shared_infra": "Shared infrastructure",
+    "shared_infra_blurb": (
+        "Other endpoints with the same fingerprint (a hint, not proof of the same operator):"
+    ),
+    "recent_history": "Recent history",
+    "when": "When (UTC)",
+    "valid": "Valid",
+    "error": "Error",
+    "yes": "yes",
+    "no": "no",
+    "no_probes": "No probes yet",
+    "not_found": "Tracker not found",
+    "bootstrap_empty": (
+        "Empty inventory: could not import sources on startup ({exc}). "
+        "The site is up; retry with `python -m app.cli import-sources`."
+    ),
+    "bootstrap_stale": (
+        "Stale lists: could not import sources on startup ({exc}). "
+        "The site is up; retry with `python -m app.cli import-sources`."
+    ),
+}
+
+STRINGS = {"pt": _PT, "en": _EN}
+
+_BOOTSTRAP_KEYS = {
+    "empty_import": "bootstrap_empty",
+    "stale_import": "bootstrap_stale",
+}
+
+
+def normalize_lang(code: str | None) -> str:
+    if not code:
+        return DEFAULT
+    tag = code.strip().lower().replace("_", "-")
+    if tag in LOCALES:
+        return tag
+    if tag.startswith("en"):
+        return "en"
+    if tag.startswith("pt"):
+        return "pt"
+    return DEFAULT
+
+
+def from_accept_language(header: str | None) -> str:
+    if not header:
+        return DEFAULT
+    for part in header.split(","):
+        tag = part.split(";", 1)[0].strip()
+        if not tag:
+            continue
+        return normalize_lang(tag)
+    return DEFAULT
+
+
+def resolve_lang(request: Request) -> str:
+    cookie = request.cookies.get(COOKIE)
+    if cookie in LOCALES:
+        return cookie
+    return from_accept_language(request.headers.get("accept-language"))
+
+
+def html_lang(lang: str) -> str:
+    return "pt-BR" if lang == "pt" else "en"
+
+
+def translate(lang: str, key: str, **kwargs: object) -> str:
+    table = STRINGS.get(lang) or STRINGS[DEFAULT]
+    text = table.get(key) or STRINGS[DEFAULT].get(key) or key
+    if kwargs:
+        return text.format(**kwargs)
+    return text
+
+
+def translator(lang: str):
+    def _t(key: str, **kwargs: object) -> str:
+        return translate(lang, key, **kwargs)
+
+    return _t
+
+
+def encode_bootstrap_note(kind: str, exc: object) -> str:
+    return f"{kind}|{exc}"
+
+
+def render_bootstrap_note(lang: str, raw: str | None) -> str:
+    if not raw:
+        return ""
+    kind, sep, detail = raw.partition("|")
+    key = _BOOTSTRAP_KEYS.get(kind)
+    if key and sep:
+        return translate(lang, key, exc=detail)
+    return raw
+
+
+def safe_back(request: Request) -> str:
+    ref = request.headers.get("referer")
+    if not ref:
+        return "/"
+    parsed = urlparse(ref)
+    if parsed.netloc and parsed.netloc != request.url.netloc:
+        return "/"
+    path = parsed.path or "/"
+    if path.startswith("/lang"):
+        return "/"
+    if parsed.query:
+        return f"{path}?{parsed.query}"
+    return path
+
+
+def lang_redirect(request: Request, code: str) -> RedirectResponse:
+    lang = normalize_lang(code)
+    resp = RedirectResponse(safe_back(request), status_code=303)
+    resp.set_cookie(
+        COOKIE,
+        lang,
+        max_age=COOKIE_MAX_AGE,
+        path="/",
+        samesite="lax",
+        httponly=True,
+    )
+    return resp

@@ -0,0 +1,128 @@
+from __future__ import annotations
+
+import logging
+from contextlib import asynccontextmanager
+from datetime import UTC, datetime, timedelta
+from pathlib import Path
+
+from fastapi import FastAPI
+from fastapi.responses import PlainTextResponse
+from fastapi.staticfiles import StaticFiles
+from sqlalchemy import text
+
+from app.api import pages, routes
+from app.collectors import import_all_sources
+from app.config import get_settings
+from app.db import get_engine, get_session_factory, init_db
+from app.db.queries import count_trackers, latest_successful_import_at, set_app_state
+from app.i18n import encode_bootstrap_note
+from app.scheduler import start_scheduler, stop_scheduler
+
+logger = logging.getLogger(__name__)
+
+
+def _configure_logging(level: str) -> None:
+    # force=True: uvicorn configures logging before lifespan runs, so without it
+    # basicConfig is a no-op and app logs (scheduler, imports) never reach stdout.
+    logging.basicConfig(
+        level=getattr(logging, level.upper(), logging.INFO),
+        format="%(asctime)s %(levelname)s [%(name)s] %(message)s",
+        force=True,
+    )
+
+
+async def _run_migrations() -> None:
+    """Apply Alembic migrations; fall back to create_all if needed."""
+    try:
+        from alembic import command
+        from alembic.config import Config
+
+        cfg = Config(str(Path(__file__).resolve().parents[1] / "alembic.ini"))
+        settings = get_settings()
+        cfg.set_main_option("sqlalchemy.url", settings.database_url)
+        await __import__("asyncio").to_thread(command.upgrade, cfg, "head")
+    except Exception:  # noqa: BLE001
+        logger.exception("Alembic upgrade failed; falling back to create_all")
+        await init_db()
+
+
+@asynccontextmanager
+async def lifespan(app: FastAPI):
+    settings = get_settings()
+    _configure_logging(settings.log_level)
+    get_engine(settings)
+    await _run_migrations()
+
+    factory = get_session_factory()
+    async with factory() as session:
+        total = await count_trackers(session)
+        last_import = await latest_successful_import_at(session)
+        max_age = timedelta(hours=settings.import_max_age_hours)
+        if last_import is not None and last_import.tzinfo is None:
+            last_import = last_import.replace(tzinfo=UTC)
+        stale = last_import is None or (datetime.now(UTC) - last_import) > max_age
+        if total == 0 or stale:
+            try:
+                await import_all_sources(session, settings)
+                await set_app_state(session, "bootstrap_note", "")
+            except Exception as exc:  # noqa: BLE001
+                kind = "empty_import" if total == 0 else "stale_import"
+                logger.warning("bootstrap import failed (%s): %s", kind, exc)
+                await set_app_state(session, "bootstrap_note", encode_bootstrap_note(kind, exc))
+
+    if settings.run_scheduler:
+        start_scheduler(settings)
+    yield
+    stop_scheduler()
+
+
+def create_app() -> FastAPI:
+    settings = get_settings()
+    app = FastAPI(
+        title=settings.app_name,
+        description=settings.app_description,
+        version=settings.app_version,
+        lifespan=lifespan,
+    )
+
+    static_dir = Path(__file__).resolve().parent / "static"
+    app.mount("/static", StaticFiles(directory=str(static_dir)), name="static")
+
+    app.include_router(pages.router)
+    app.include_router(routes.router, prefix="/api/v1")
+
+    @app.get("/healthz", response_class=PlainTextResponse)
+    async def healthz() -> str:
+        return "ok"
+
+    @app.get("/readyz", response_class=PlainTextResponse)
+    async def readyz() -> PlainTextResponse:
+        try:
+            engine = get_engine()
+            async with engine.connect() as conn:
+                await conn.execute(text("SELECT 1"))
+            return PlainTextResponse("ready")
+        except Exception:  # noqa: BLE001
+            return PlainTextResponse("not ready", status_code=503)
+
+    return app
+
+
+app = create_app()
+
+
+def run() -> None:
+    import uvicorn
+
+    settings = get_settings()
+    uvicorn.run(
+        "app.main:app",
+        host=settings.host,
+        port=settings.port,
+        workers=1,
+        log_level=settings.log_level.lower(),
+    )
+
+
+if __name__ == "__main__":
+    run()

@@ -0,0 +1,209 @@
+from __future__ import annotations
+
+import asyncio
+import logging
+import random
+from datetime import UTC, datetime, timedelta
+
+import httpx
+from sqlalchemy import select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.config import Settings
+from app.db.models import DnsRecord, ProbeResult, Tracker
+from app.probes.http import probe_http
+from app.probes.udp import probe_udp
+from app.services.dns import host_has_ipv6, lookup_asn, resolve_cname_chain
+
+logger = logging.getLogger(__name__)
+
+
+def compute_next_check(
+    *,
+    settings: Settings,
+    announced_interval: int | None,
+    consecutive_failures: int,
+    success: bool,
+) -> datetime:
+    now = datetime.now(UTC)
+    if success:
+        base = announced_interval or settings.default_probe_interval_seconds
+        base = max(base, settings.min_probe_interval_seconds)
+    else:
+        exp = min(consecutive_failures, 12)
+        base = min(
+            settings.min_probe_interval_seconds * (2 ** max(exp - 1, 0)),
+            12 * 3600,
+        )
+        base = max(base, settings.min_probe_interval_seconds)
+    jitter = random.uniform(0, min(60.0, base * 0.1))
+    return now + timedelta(seconds=base + jitter)
+
+
+async def _upsert_dns_records(
+    session: AsyncSession, tracker: Tracker, records: list[tuple[str, str]]
+) -> None:
+    now = datetime.now(UTC)
+    for rtype, value in records:
+        result = await session.execute(
+            select(DnsRecord).where(
+                DnsRecord.tracker_id == tracker.id,
+                DnsRecord.record_type == rtype,
+                DnsRecord.value == value,
+            )
+        )
+        row = result.scalar_one_or_none()
+        if row is None:
+            session.add(
+                DnsRecord(
+                    tracker_id=tracker.id,
+                    record_type=rtype,
+                    value=value,
+                    first_seen_at=now,
+                    last_seen_at=now,
+                )
+            )
+        else:
+            row.last_seen_at = now
+
+
+async def probe_tracker(
+    session: AsyncSession,
+    tracker: Tracker,
+    settings: Settings,
+    client: httpx.AsyncClient,
+) -> ProbeResult:
+    now = datetime.now(UTC)
+    dns = await resolve_cname_chain(tracker.hostname)
+    tracker.terminal_cname = dns.terminal_cname
+    tracker.infrastructure_fingerprint = dns.fingerprint
+
+    dns_rows: list[tuple[str, str]] = [("CNAME", c) for c in dns.cname_chain]
+    dns_rows.extend(("A", ip) for ip in dns.ipv4)
+    dns_rows.extend(("AAAA", ip) for ip in dns.ipv6)
+    await _upsert_dns_records(session, tracker, dns_rows)
+
+    ipv6_mode = settings.ipv6_mode()
+    can_v6 = host_has_ipv6() if ipv6_mode == "auto" else ipv6_mode == "true"
+    tracker.supports_ipv6 = "not_tested" if not can_v6 else ("true" if dns.ipv6 else "false")
+    tracker.supports_ipv4 = bool(dns.ipv4) if settings.enable_ipv4 else None
+
+    public_ips = list(dns.ipv4)
+    if can_v6:
+        public_ips.extend(dns.ipv6)
+
+    if settings.geoip_asn_db and public_ips:
+        asn, net, country = lookup_asn(public_ips[0], settings.geoip_asn_db)
+        tracker.asn = asn
+        tracker.network_name = net
+        tracker.country_code = country
+
+    if tracker.scheme == "udp":
+        peer = (settings.peer_id_prefix() + b"xxxxxxxxxxxx")[:20]
+        outcome = await probe_udp(
+            tracker.hostname,
+            tracker.port,
+            timeout=settings.probe_timeout_seconds,
+            peer_id=peer,
+        )
+    elif tracker.scheme in {"http", "https"}:
+        outcome = await probe_http(
+            tracker.canonical_url,
+            client=client,
+            timeout=settings.probe_timeout_seconds,
+            max_bytes=settings.max_response_bytes,
+            peer_id_prefix=settings.peer_id_prefix(),
+            resolved_ips=public_ips or None,
+        )
+    else:
+        tracker.current_status = "unsupported"
+        tracker.last_checked_at = now
+        tracker.next_check_at = now + timedelta(days=30)
+        result = ProbeResult(
+            tracker_id=tracker.id,
+            checked_at=now,
+            status="unsupported",
+            response_valid=False,
+            error_kind="unsupported",
+            error_detail="scheme not probed",
+        )
+        session.add(result)
+        await session.commit()
+        return result
+
+    success = outcome.response_valid and outcome.status == "up"
+    if success:
+        tracker.consecutive_failures = 0
+        tracker.current_status = "up"
+        if outcome.tracker_interval_seconds:
+            tracker.announced_interval_seconds = outcome.tracker_interval_seconds
+    else:
+        tracker.consecutive_failures += 1
+        tracker.current_status = (
+            outcome.status if outcome.status in {"down", "degraded"} else "down"
+        )
+
+    tracker.last_latency_ms = outcome.latency_ms
+    tracker.last_checked_at = now
+    tracker.next_check_at = compute_next_check(
+        settings=settings,
+        announced_interval=tracker.announced_interval_seconds,
+        consecutive_failures=tracker.consecutive_failures,
+        success=success,
+    )
+
+    result = ProbeResult(
+        tracker_id=tracker.id,
+        checked_at=now,
+        status=tracker.current_status,
+        latency_ms=outcome.latency_ms,
+        ip_address=outcome.ip_address,
+        ip_family=outcome.ip_family,
+        response_valid=outcome.response_valid,
+        tracker_interval_seconds=outcome.tracker_interval_seconds,
+        seeders=outcome.seeders,
+        leechers=outcome.leechers,
+        error_kind=outcome.error_kind,
+        error_detail=(outcome.error_detail or "")[:512] or None,
+    )
+    session.add(result)
+    await session.commit()
+    return result
+
+
+async def probe_due_batch(
+    session: AsyncSession,
+    settings: Settings,
+    *,
+    limit: int | None = None,
+) -> int:
+    from app.db import get_session_factory
+    from app.db.queries import due_trackers
+
+    limit = limit or settings.probe_concurrency
+    trackers = await due_trackers(session, limit=limit)
+    if not trackers:
+        return 0
+
+    ids = [t.id for t in trackers]
+    sem = asyncio.Semaphore(settings.probe_concurrency)
+    factory = get_session_factory()
+
+    async with httpx.AsyncClient(
+        timeout=settings.probe_timeout_seconds,
+        headers={"User-Agent": settings.user_agent()},
+        follow_redirects=False,
+    ) as client:
+
+        async def _one(tracker_id: int) -> None:
+            async with sem, factory() as own_session:
+                tracker = await own_session.get(Tracker, tracker_id)
+                if tracker is None:
+                    return
+                try:
+                    await probe_tracker(own_session, tracker, settings, client)
+                except Exception:  # noqa: BLE001
+                    logger.exception("probe failed for id=%s", tracker_id)
+
+        await asyncio.gather(*[_one(i) for i in ids])
+    return len(ids)

@@ -0,0 +1,156 @@
+"""Bencode decoder for tracker announce responses.
+
+Adapted from newTrackon (MIT). Rewritten for this architecture.
+Peers are validated structurally and discarded by callers; this module never
+persists peer addresses.
+"""
+
+from __future__ import annotations
+
+from collections import OrderedDict
+from socket import AF_INET, AF_INET6, inet_ntop
+from struct import unpack_from
+from typing import Any
+
+TOK_DICT = b"d"
+TOK_LIST = b"l"
+TOK_INT = b"i"
+TOK_END = b"e"
+TOK_STR_SEP = b":"
+
+BDecoded = OrderedDict[bytes, Any] | list[Any] | int | bytes | None
+
+
+class BencodeError(ValueError):
+    pass
+
+
+def bdecode(data: bytes) -> dict[str, Any]:
+    """Decode a bencoded tracker response into a string-keyed dict.
+
+    Binary `peers` / `peers6` are converted to length-checked peer lists for
+    validation, then callers must discard addresses immediately.
+    """
+    if not data:
+        raise BencodeError("empty bencode payload")
+    decoded = Decoder(data).decode()
+    if not isinstance(decoded, OrderedDict):
+        raise BencodeError("tracker response must be a dict")
+    response: dict[str, Any] = {}
+    for key, value in decoded.items():
+        response[key.decode("utf-8", errors="replace")] = value
+
+    if "peers" in response and isinstance(response["peers"], bytes):
+        response["peers"] = decode_binary_peers_list(response["peers"], 0, AF_INET)
+    if "peers6" in response and isinstance(response["peers6"], bytes):
+        response["peers6"] = decode_binary_peers_list(response["peers6"], 0, AF_INET6)
+
+    for key, value in list(response.items()):
+        if isinstance(value, bytes):
+            response[key] = value.decode("utf-8", errors="replace")
+    return response
+
+
+def decode_binary_peers_list(buf: bytes, offset: int, ip_family: int) -> list[dict[str, Any]]:
+    """Validate compact peer binary layout; return ephemeral peer dicts."""
+    peers: list[dict[str, Any]] = []
+    peer_length = 6 if ip_family == AF_INET else 18
+    view = memoryview(buf)
+    while offset < len(buf):
+        if len(buf) < offset + peer_length:
+            break
+        ip_bytes = bytes(view[offset : offset + peer_length - 2])
+        try:
+            ip_str = inet_ntop(ip_family, ip_bytes)
+        except OSError as exc:
+            raise BencodeError("invalid compact peer address") from exc
+        offset += peer_length - 2
+        port = unpack_from("!H", buf, offset)[0]
+        offset += 2
+        peers.append({"ip": ip_str, "port": port})
+    return peers
+
+
+def validate_peers_then_discard(response: dict[str, Any]) -> bool:
+    """Return True if peers/peers6 field exists and is structurally valid, then clear it."""
+    has_peers = False
+    for key in ("peers", "peers6"):
+        if key not in response:
+            continue
+        value = response[key]
+        if isinstance(value, list):
+            has_peers = True
+        elif isinstance(value, (bytes, str)):
+            # dictionary-model peers — accept presence without storing
+            has_peers = True
+        response[key] = None
+    return has_peers
+
+
+class Decoder:
+    def __init__(self, data: bytes) -> None:
+        self.index = 0
+        self.data = data
+
+    def decode(self) -> BDecoded:
+        c = self.peek()
+        if c is None:
+            raise BencodeError("unexpected EOF")
+        if c == TOK_DICT:
+            self.read(1)
+            return self.decode_dict()
+        if c == TOK_LIST:
+            self.read(1)
+            return self.decode_list()
+        if c == TOK_INT:
+            self.read(1)
+            return self.decode_int()
+        if c in b"0123456789":
+            return self.decode_str()
+        raise BencodeError("invalid bencode token")
+
+    def peek(self) -> bytes | None:
+        if self.index >= len(self.data):
+            return None
+        return self.data[self.index : self.index + 1]
+
+    def read(self, length: int) -> bytes:
+        if self.index + length > len(self.data):
+            raise BencodeError("unexpected EOF while reading")
+        result = self.data[self.index : self.index + length]
+        self.index += length
+        return result
+
+    def read_until(self, token: bytes) -> bytes:
+        loc = self.data.find(token, self.index)
+        if loc == -1:
+            raise BencodeError("token not found")
+        result = self.data[self.index : loc]
+        self.index = loc + 1
+        return result
+
+    def decode_dict(self) -> OrderedDict[bytes, BDecoded]:
+        result: OrderedDict[bytes, BDecoded] = OrderedDict()
+        while self.data[self.index : self.index + 1] != TOK_END:
+            key = self.decode()
+            if not isinstance(key, bytes):
+                raise BencodeError("dict key must be bytes")
+            result[key] = self.decode()
+        self.read(1)
+        return result
+
+    def decode_list(self) -> list[BDecoded]:
+        result: list[BDecoded] = []
+        while self.data[self.index : self.index + 1] != TOK_END:
+            result.append(self.decode())
+        self.read(1)
+        return result
+
+    def decode_int(self) -> int:
+        return int(self.read_until(TOK_END))
+
+    def decode_str(self) -> bytes:
+        length = int(self.read_until(TOK_STR_SEP))
+        if length < 0:
+            raise BencodeError("negative string length")
+        return self.read(length)

@@ -0,0 +1,169 @@
+from __future__ import annotations
+
+import os
+import time
+
+import httpx
+
+from app.probes.bencode import BencodeError, bdecode, validate_peers_then_discard
+from app.probes.types import ProbeOutcome
+from app.services.ssrf import assert_safe_destination, is_public_ip
+
+
+async def probe_http(
+    url: str,
+    *,
+    client: httpx.AsyncClient,
+    timeout: float = 10.0,
+    max_bytes: int = 1_048_576,
+    peer_id_prefix: bytes = b"-RASTRO-",
+    info_hash: bytes | None = None,
+    resolved_ips: list[str] | None = None,
+) -> ProbeOutcome:
+    info_hash = info_hash or os.urandom(20)
+    peer_id = (peer_id_prefix[:8] + os.urandom(12))[:20]
+    from urllib.parse import quote_from_bytes
+
+    query = (
+        f"info_hash={quote_from_bytes(info_hash)}"
+        f"&peer_id={quote_from_bytes(peer_id)}"
+        f"&port=0&uploaded=0&downloaded=0&left=0&compact=1&numwant=0&event=started"
+    )
+    announce_url = f"{url}?{query}" if "?" not in url else f"{url}&{query}"
+
+    t0 = time.perf_counter()
+    try:
+        if resolved_ips is not None:
+            host = httpx.URL(url).host or ""
+            assert_safe_destination(host, resolved_ips)
+
+        async with client.stream(
+            "GET",
+            announce_url,
+            timeout=timeout,
+            follow_redirects=False,
+        ) as response:
+            # Re-validate resolved addresses before consuming the body (anti-rebinding).
+            if resolved_ips:
+                for ip in resolved_ips:
+                    if not is_public_ip(ip):
+                        return ProbeOutcome(
+                            status="down",
+                            response_valid=False,
+                            error_kind="blocked",
+                            error_detail=f"blocked address: {ip}",
+                            latency_ms=(time.perf_counter() - t0) * 1000,
+                        )
+
+            chunks: list[bytes] = []
+            total = 0
+            async for chunk in response.aiter_bytes():
+                total += len(chunk)
+                if total > max_bytes:
+                    return ProbeOutcome(
+                        status="down",
+                        response_valid=False,
+                        error_kind="too_large",
+                        error_detail="response exceeded MAX_RESPONSE_BYTES",
+                        latency_ms=(time.perf_counter() - t0) * 1000,
+                    )
+                chunks.append(chunk)
+            body = b"".join(chunks)
+            latency = (time.perf_counter() - t0) * 1000
+
+            if response.status_code != 200:
+                return ProbeOutcome(
+                    status="down",
+                    response_valid=False,
+                    latency_ms=latency,
+                    error_kind="http_status",
+                    error_detail=f"HTTP {response.status_code}",
+                )
+            if not body:
+                return ProbeOutcome(
+                    status="down",
+                    response_valid=False,
+                    latency_ms=latency,
+                    error_kind="empty",
+                    error_detail="empty HTTP response",
+                )
+            try:
+                decoded = bdecode(body)
+            except BencodeError as exc:
+                return ProbeOutcome(
+                    status="down",
+                    response_valid=False,
+                    latency_ms=latency,
+                    error_kind="malformed",
+                    error_detail=str(exc)[:512],
+                )
+
+            if "failure reason" in decoded:
+                return ProbeOutcome(
+                    status="degraded",
+                    response_valid=False,
+                    latency_ms=latency,
+                    error_kind="tracker_error",
+                    error_detail=str(decoded["failure reason"])[:512],
+                )
+
+            # Require peers field OR interval for a minimal healthy announce
+            has_peers = validate_peers_then_discard(decoded)
+            interval = decoded.get("interval")
+            if not has_peers and interval is None:
+                return ProbeOutcome(
+                    status="degraded",
+                    response_valid=False,
+                    latency_ms=latency,
+                    error_kind="incomplete",
+                    error_detail="missing peers and interval",
+                )
+
+            seeders = decoded.get("complete")
+            leechers = decoded.get("incomplete")
+            return ProbeOutcome(
+                status="up",
+                response_valid=True,
+                latency_ms=latency,
+                ip_address=resolved_ips[0] if resolved_ips else None,
+                ip_family="4"
+                if resolved_ips and ":" not in resolved_ips[0]
+                else ("6" if resolved_ips else None),
+                tracker_interval_seconds=int(interval) if isinstance(interval, int) else None,
+                seeders=int(seeders) if isinstance(seeders, int) else None,
+                leechers=int(leechers) if isinstance(leechers, int) else None,
+            )
+    except httpx.TimeoutException:
+        return ProbeOutcome(
+            status="down",
+            response_valid=False,
+            error_kind="timeout",
+            error_detail="HTTP timeout",
+            latency_ms=(time.perf_counter() - t0) * 1000,
+        )
+    except httpx.ConnectError as exc:
+        detail = str(exc)
+        kind = "tls" if "SSL" in detail.upper() or "TLS" in detail.upper() else "connection"
+        return ProbeOutcome(
+            status="down",
+            response_valid=False,
+            error_kind=kind,
+            error_detail=detail[:512],
+            latency_ms=(time.perf_counter() - t0) * 1000,
+        )
+    except ValueError as exc:
+        return ProbeOutcome(
+            status="down",
+            response_valid=False,
+            error_kind="blocked",
+            error_detail=str(exc)[:512],
+            latency_ms=(time.perf_counter() - t0) * 1000,
+        )
+    except httpx.HTTPError as exc:
+        return ProbeOutcome(
+            status="down",
+            response_valid=False,
+            error_kind="http",
+            error_detail=str(exc)[:512],
+            latency_ms=(time.perf_counter() - t0) * 1000,
+        )

@@ -0,0 +1,17 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+
+
+@dataclass(slots=True)
+class ProbeOutcome:
+    status: str  # up | down | degraded
+    response_valid: bool
+    latency_ms: float | None = None
+    ip_address: str | None = None
+    ip_family: str | None = None
+    tracker_interval_seconds: int | None = None
+    seeders: int | None = None
+    leechers: int | None = None
+    error_kind: str | None = None
+    error_detail: str | None = None

@@ -0,0 +1,170 @@
+"""UDP BitTorrent tracker probe (BEP 15).
+
+Protocol packing inspired by newTrackon (MIT).
+"""
+
+from __future__ import annotations
+
+import asyncio
+import os
+import random
+import socket
+import struct
+import time
+
+from app.probes.types import ProbeOutcome
+from app.services.ssrf import assert_connected_ip_safe, assert_safe_destination
+
+
+def _transaction_id() -> int:
+    return random.randint(0, 0x7FFFFFFF)
+
+
+def create_connect_request() -> tuple[bytes, int]:
+    connection_id = 0x41727101980
+    action = 0
+    tid = _transaction_id()
+    buf = struct.pack("!qii", connection_id, action, tid)
+    return buf, tid
+
+
+def parse_connect_response(buf: bytes, sent_tid: int) -> int:
+    if len(buf) < 16:
+        raise RuntimeError(f"connect response too short: {len(buf)}")
+    action, tid = struct.unpack_from("!ii", buf, 0)
+    if tid != sent_tid:
+        raise RuntimeError("transaction id mismatch on connect")
+    if action == 3:
+        raise RuntimeError("tracker returned UDP error on connect")
+    if action != 0:
+        raise RuntimeError(f"unexpected connect action: {action}")
+    return struct.unpack_from("!q", buf, 8)[0]
+
+
+def create_announce_request(
+    connection_id: int,
+    info_hash: bytes,
+    peer_id: bytes,
+    *,
+    num_want: int = 0,
+    port: int = 0,
+) -> tuple[bytes, int]:
+    action = 1
+    tid = _transaction_id()
+    buf = struct.pack("!qii", connection_id, action, tid)
+    buf += struct.pack("!20s20s", info_hash, peer_id)
+    buf += struct.pack("!qqq", 0, 0, 0)  # downloaded, left, uploaded
+    buf += struct.pack("!i", 2)  # event = started
+    buf += struct.pack("!i", 0)  # IP
+    buf += struct.pack("!i", _transaction_id())  # key
+    buf += struct.pack("!i", num_want)
+    buf += struct.pack("!H", port)
+    return buf, tid
+
+
+def parse_announce_response(buf: bytes, sent_tid: int) -> tuple[int, int, int]:
+    """Return (interval, leechers, seeders). Peer bytes are ignored/discarded."""
+    if len(buf) < 20:
+        raise RuntimeError(f"announce response too short: {len(buf)}")
+    action, tid = struct.unpack_from("!ii", buf, 0)
+    if tid != sent_tid:
+        raise RuntimeError("transaction id mismatch on announce")
+    if action == 3:
+        raise RuntimeError("tracker returned UDP error on announce")
+    if action != 1:
+        raise RuntimeError(f"unexpected announce action: {action}")
+    interval, leechers, seeders = struct.unpack_from("!iii", buf, 8)
+    # Remaining bytes would be peers — intentionally not parsed into storage.
+    return interval, leechers, seeders
+
+
+async def probe_udp(
+    host: str,
+    port: int,
+    *,
+    timeout: float = 10.0,
+    info_hash: bytes | None = None,
+    peer_id: bytes | None = None,
+) -> ProbeOutcome:
+    info_hash = info_hash or os.urandom(20)
+    peer_id = peer_id or (b"-RASTRO-" + os.urandom(12))
+    if len(peer_id) != 20:
+        peer_id = (peer_id + os.urandom(20))[:20]
+
+    loop = asyncio.get_running_loop()
+    t0 = time.perf_counter()
+    try:
+        infos = await loop.getaddrinfo(host, port, type=socket.SOCK_DGRAM)
+    except OSError as exc:
+        return ProbeOutcome(
+            status="down",
+            response_valid=False,
+            error_kind="dns",
+            error_detail=str(exc)[:512],
+            latency_ms=(time.perf_counter() - t0) * 1000,
+        )
+
+    ips = [str(item[4][0]) for item in infos]
+    try:
+        assert_safe_destination(host, ips)
+    except ValueError as exc:
+        return ProbeOutcome(
+            status="down",
+            response_valid=False,
+            error_kind="blocked",
+            error_detail=str(exc)[:512],
+            latency_ms=(time.perf_counter() - t0) * 1000,
+        )
+
+    last_error = "udp announce failed"
+    for af, socktype, proto, _, sockaddr in infos:
+        ip = str(sockaddr[0])
+        try:
+            assert_connected_ip_safe(ip)
+        except ValueError as exc:
+            last_error = str(exc)
+            continue
+
+        sock = socket.socket(af, socktype, proto)
+        sock.setblocking(False)
+        try:
+            await asyncio.wait_for(loop.sock_connect(sock, sockaddr), timeout=timeout)
+            connect_req, connect_tid = create_connect_request()
+            await asyncio.wait_for(loop.sock_sendall(sock, connect_req), timeout=timeout)
+            connect_buf = await asyncio.wait_for(loop.sock_recv(sock, 2048), timeout=timeout)
+            connection_id = parse_connect_response(connect_buf, connect_tid)
+
+            announce_req, announce_tid = create_announce_request(
+                connection_id, info_hash, peer_id, num_want=0
+            )
+            await asyncio.wait_for(loop.sock_sendall(sock, announce_req), timeout=timeout)
+            announce_buf = await asyncio.wait_for(loop.sock_recv(sock, 2048), timeout=timeout)
+            interval, leechers, seeders = parse_announce_response(announce_buf, announce_tid)
+            latency = (time.perf_counter() - t0) * 1000
+            family = "4" if af == socket.AF_INET else "6"
+            return ProbeOutcome(
+                status="up",
+                response_valid=True,
+                latency_ms=latency,
+                ip_address=ip,
+                ip_family=family,
+                tracker_interval_seconds=interval,
+                seeders=seeders,
+                leechers=leechers,
+            )
+        except TimeoutError:
+            last_error = "udp timeout"
+        except OSError as exc:
+            last_error = f"udp error: {exc}"
+        except RuntimeError as exc:
+            last_error = str(exc)
+        finally:
+            sock.close()
+
+    return ProbeOutcome(
+        status="down",
+        response_valid=False,
+        error_kind="udp",
+        error_detail=last_error[:512],
+        latency_ms=(time.perf_counter() - t0) * 1000,
+    )

@@ -0,0 +1,100 @@
+from __future__ import annotations
+
+import asyncio
+import logging
+
+from apscheduler.schedulers.asyncio import AsyncIOScheduler
+
+from app.collectors import import_all_sources
+from app.config import Settings, get_settings
+from app.db import get_session_factory
+from app.db.queries import prune_probe_results
+from app.probes import probe_due_batch
+
+logger = logging.getLogger(__name__)
+
+_scheduler: AsyncIOScheduler | None = None
+_locks = {
+    "import": asyncio.Lock(),
+    "probe": asyncio.Lock(),
+    "prune": asyncio.Lock(),
+}
+
+
+async def job_import_sources() -> None:
+    if _locks["import"].locked():
+        logger.warning("import job already running; skipping")
+        return
+    async with _locks["import"]:
+        settings = get_settings()
+        factory = get_session_factory()
+        async with factory() as session:
+            await import_all_sources(session, settings)
+
+
+async def job_probe_due() -> None:
+    if _locks["probe"].locked():
+        logger.warning("probe job already running; skipping")
+        return
+    async with _locks["probe"]:
+        settings = get_settings()
+        factory = get_session_factory()
+        async with factory() as session:
+            count = await probe_due_batch(session, settings)
+            logger.info("probed %s trackers", count)
+
+
+async def job_prune() -> None:
+    if _locks["prune"].locked():
+        return
+    async with _locks["prune"]:
+        settings = get_settings()
+        factory = get_session_factory()
+        async with factory() as session:
+            removed = await prune_probe_results(session, settings.raw_result_retention_days)
+            logger.info("pruned %s probe results", removed)
+
+
+def start_scheduler(settings: Settings | None = None) -> AsyncIOScheduler:
+    global _scheduler
+    settings = settings or get_settings()
+    if _scheduler is not None:
+        return _scheduler
+    scheduler = AsyncIOScheduler(timezone="UTC")
+    scheduler.add_job(
+        job_import_sources,
+        "cron",
+        day_of_week=settings.import_day_of_week,
+        hour=settings.import_hour_utc,
+        minute=0,
+        id="import_sources",
+        max_instances=1,
+        coalesce=True,
+    )
+    scheduler.add_job(
+        job_probe_due,
+        "interval",
+        seconds=60,
+        id="probe_due",
+        max_instances=1,
+        coalesce=True,
+    )
+    scheduler.add_job(
+        job_prune,
+        "interval",
+        hours=24,
+        id="prune_results",
+        max_instances=1,
+        coalesce=True,
+    )
+    scheduler.start()
+    _scheduler = scheduler
+    logger.info("scheduler started (single process / single worker required)")
+    return scheduler
+
+
+def stop_scheduler() -> None:
+    global _scheduler
+    if _scheduler is not None:
+        _scheduler.shutdown(wait=False)
+        _scheduler = None

@@ -0,0 +1,123 @@
+from __future__ import annotations
+
+import hashlib
+import logging
+import socket
+from dataclasses import dataclass, field
+
+import dns.asyncresolver
+
+from app.services.ssrf import is_public_ip
+
+logger = logging.getLogger(__name__)
+
+MAX_CNAME_DEPTH = 10
+
+
+@dataclass
+class DnsSnapshot:
+    cname_chain: list[str] = field(default_factory=list)
+    terminal_cname: str | None = None
+    ipv4: list[str] = field(default_factory=list)
+    ipv6: list[str] = field(default_factory=list)
+    cycle_detected: bool = False
+    fingerprint: str | None = None
+
+
+async def resolve_cname_chain(hostname: str, *, max_depth: int = MAX_CNAME_DEPTH) -> DnsSnapshot:
+    snapshot = DnsSnapshot()
+    resolver = dns.asyncresolver.Resolver()
+    current = hostname.rstrip(".").lower()
+    seen: set[str] = set()
+
+    for _ in range(max_depth):
+        if current in seen:
+            snapshot.cycle_detected = True
+            break
+        seen.add(current)
+        try:
+            answer = await resolver.resolve(current, "CNAME")
+        except Exception:  # noqa: BLE001 — end of CNAME chain
+            break
+        if not answer:
+            break
+        target = str(answer[0].target).rstrip(".").lower()
+        snapshot.cname_chain.append(target)
+        current = target
+
+    if snapshot.cname_chain:
+        snapshot.terminal_cname = snapshot.cname_chain[-1]
+
+    lookup_host = snapshot.terminal_cname or hostname.rstrip(".").lower()
+    snapshot.ipv4 = await _resolve_addresses(resolver, lookup_host, "A")
+    snapshot.ipv6 = await _resolve_addresses(resolver, lookup_host, "AAAA")
+    snapshot.fingerprint = compute_fingerprint(
+        snapshot.terminal_cname, snapshot.ipv4 + snapshot.ipv6
+    )
+    return snapshot
+
+
+async def _resolve_addresses(
+    resolver: dns.asyncresolver.Resolver, hostname: str, rdtype: str
+) -> list[str]:
+    try:
+        answer = await resolver.resolve(hostname, rdtype)
+    except Exception:  # noqa: BLE001
+        return []
+    out: list[str] = []
+    for rdata in answer:
+        ip = rdata.to_text()
+        if is_public_ip(ip):
+            out.append(ip)
+    return sorted(set(out))
+
+
+def compute_fingerprint(terminal_cname: str | None, public_ips: list[str]) -> str | None:
+    if terminal_cname:
+        material = f"cname:{terminal_cname.lower()}"
+    elif public_ips:
+        material = "ips:" + ",".join(sorted(set(public_ips)))
+    else:
+        return None
+    return hashlib.sha256(material.encode("utf-8")).hexdigest()[:32]
+
+
+def host_has_ipv6() -> bool:
+    """Best-effort detection of local IPv6 connectivity."""
+    try:
+        sock = socket.socket(socket.AF_INET6, socket.SOCK_DGRAM)
+        try:
+            sock.connect(("2001:4860:4860::8888", 53))
+            return True
+        finally:
+            sock.close()
+    except OSError:
+        return False
+
+
+def lookup_asn(ip: str, db_path: str) -> tuple[int | None, str | None, str | None]:
+    """Optional MaxMind ASN lookup. Returns (asn, network_name, country_code)."""
+    if not db_path:
+        return None, None, None
+    try:
+        import geoip2.database  # type: ignore
+    except ImportError:
+        return None, None, None
+    try:
+        with geoip2.database.Reader(db_path) as reader:
+            # Prefer ASN db; country may be unavailable in ASN-only files
+            try:
+                asn_resp = reader.asn(ip)
+                asn = asn_resp.autonomous_system_number
+                org = asn_resp.autonomous_system_organization
+            except Exception:  # noqa: BLE001
+                asn, org = None, None
+            country = None
+            try:
+                country = reader.country(ip).country.iso_code  # type: ignore[attr-defined]
+            except Exception:  # noqa: BLE001
+                country = None
+            return asn, org, country
+    except Exception:  # noqa: BLE001
+        logger.debug("GeoIP lookup unavailable for %s", ip, exc_info=True)
+        return None, None, None

@@ -0,0 +1,125 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+from urllib.parse import urlparse, urlunparse
+
+SUPPORTED_SCHEMES = frozenset({"udp", "http", "https"})
+UNSUPPORTED_SCHEMES = frozenset({"ws", "wss", "i2p", "ygg"})
+
+
+class NormalizationError(ValueError):
+    """Raised when a URL cannot be normalized into a supported tracker endpoint."""
+
+
+@dataclass(frozen=True, slots=True)
+class NormalizedTracker:
+    canonical_url: str
+    scheme: str
+    hostname: str
+    port: int
+    path: str
+    unsupported: bool = False
+
+
+def _default_port(scheme: str) -> int:
+    if scheme == "https":
+        return 443
+    if scheme == "http":
+        return 80
+    return 80  # udp commonly requires explicit port; fallback unused when port missing
+
+
+def _idna_hostname(hostname: str) -> str:
+    hostname = hostname.strip().rstrip(".").lower()
+    if not hostname:
+        raise NormalizationError("missing hostname")
+    try:
+        return hostname.encode("idna").decode("ascii")
+    except UnicodeError as exc:
+        raise NormalizationError("invalid IDN hostname") from exc
+
+
+def normalize_tracker_url(raw: str) -> NormalizedTracker:
+    """Normalize a tracker URL into a canonical endpoint identity."""
+    if raw is None:
+        raise NormalizationError("empty input")
+    text = raw.strip()
+    if not text or text.startswith("#"):
+        raise NormalizationError("empty input")
+
+    # Reject credentials early even before parse edge-cases
+    if "@" in text.split("://", 1)[-1].split("/", 1)[0]:
+        raise NormalizationError("credentials not allowed")
+
+    parsed = urlparse(text)
+    scheme = (parsed.scheme or "").lower()
+    if not scheme:
+        raise NormalizationError("missing scheme")
+
+    if scheme in UNSUPPORTED_SCHEMES or scheme.endswith(".i2p") or "ygg" in scheme:
+        host = _idna_hostname(parsed.hostname or "unsupported")
+        try:
+            port = parsed.port or 0
+        except ValueError as exc:
+            raise NormalizationError("invalid port") from exc
+        return NormalizedTracker(
+            canonical_url=text,
+            scheme=scheme,
+            hostname=host,
+            port=port,
+            path=parsed.path or "/",
+            unsupported=True,
+        )
+
+    if scheme not in SUPPORTED_SCHEMES:
+        raise NormalizationError(f"unsupported scheme: {scheme}")
+
+    if parsed.username or parsed.password:
+        raise NormalizationError("credentials not allowed")
+
+    if not parsed.hostname:
+        raise NormalizationError("missing hostname")
+
+    hostname = _idna_hostname(parsed.hostname)
+
+    try:
+        port = parsed.port
+    except ValueError as exc:
+        raise NormalizationError("invalid port") from exc
+    if port is None:
+        # Prefer explicit ports when provided by source; otherwise defaults.
+        port = _default_port(scheme)
+    if not (1 <= port <= 65535):
+        raise NormalizationError("invalid port")
+
+    path = parsed.path or "/announce"
+    if not path.startswith("/"):
+        path = "/" + path
+    # Drop fragment; query is uncommon for announce base URLs — strip for identity.
+    # Preserve path as given by source (including /announce variants).
+
+    netloc = f"{hostname}:{port}"
+    canonical = urlunparse((scheme, netloc, path, "", "", ""))
+    return NormalizedTracker(
+        canonical_url=canonical,
+        scheme=scheme,
+        hostname=hostname,
+        port=port,
+        path=path,
+        unsupported=False,
+    )
+
+
+def parse_tracker_list_text(body: str) -> list[str]:
+    """Split a blank-line-tolerant tracker list into raw URL strings (deduped)."""
+    seen: set[str] = set()
+    out: list[str] = []
+    for line in body.splitlines():
+        item = line.strip()
+        if not item or item.startswith("#"):
+            continue
+        if item in seen:
+            continue
+        seen.add(item)
+        out.append(item)
+    return out

@@ -0,0 +1,151 @@
+from __future__ import annotations
+
+from collections.abc import Sequence
+from dataclasses import dataclass
+from datetime import UTC, datetime, timedelta
+from statistics import median
+
+from sqlalchemy import select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.db.models import ProbeResult, Tracker
+
+
+@dataclass(slots=True)
+class TrackerMetrics:
+    uptime_24h: float | None
+    uptime_7d: float | None
+    uptime_30d: float | None
+    latency_median_7d: float | None
+    latency_p95_7d: float | None
+    valid_rate_7d: float | None
+    measurement_count: int
+    measurement_count_7d: int
+    tracking_days: float
+    score: int | None
+    provisional: bool
+
+
+def _percentile(sorted_values: list[float], p: float) -> float | None:
+    if not sorted_values:
+        return None
+    if len(sorted_values) == 1:
+        return sorted_values[0]
+    k = (len(sorted_values) - 1) * p
+    f = int(k)
+    c = min(f + 1, len(sorted_values) - 1)
+    if f == c:
+        return sorted_values[f]
+    return sorted_values[f] + (sorted_values[c] - sorted_values[f]) * (k - f)
+
+
+def uptime_ratio(results: Sequence[ProbeResult]) -> float | None:
+    if not results:
+        return None
+    ups = sum(1 for r in results if r.status == "up" and r.response_valid)
+    return ups / len(results)
+
+
+def latency_score(median_ms: float | None) -> int:
+    if median_ms is None:
+        return 0
+    if median_ms <= 150:
+        return 20
+    if median_ms <= 300:
+        return 16
+    if median_ms <= 600:
+        return 10
+    if median_ms <= 1000:
+        return 5
+    return 0
+
+
+def confidence_score(count: int) -> int:
+    if count < 3:
+        return 0
+    if count < 10:
+        return 1
+    if count < 20:
+        return 3
+    return 5
+
+
+def compute_score(
+    *,
+    uptime_7d: float | None,
+    latency_median_7d: float | None,
+    valid_rate_7d: float | None,
+    measurement_count: int,
+) -> int | None:
+    if measurement_count < 3:
+        return None
+    avail = int(round((uptime_7d or 0.0) * 60))
+    lat = latency_score(latency_median_7d)
+    valid = int(round((valid_rate_7d or 0.0) * 15))
+    conf = confidence_score(measurement_count)
+    return max(0, min(100, avail + lat + valid + conf))
+
+
+def metrics_from_results(
+    tracker: Tracker, results: Sequence[ProbeResult], *, now: datetime | None = None
+) -> TrackerMetrics:
+    now = now or datetime.now(UTC)
+
+    # Normalize naive datetimes from SQLite
+    def _aware(dt: datetime) -> datetime:
+        return dt if dt.tzinfo else dt.replace(tzinfo=UTC)
+
+    first_seen = _aware(tracker.first_seen_at)
+    cut_24h = now - timedelta(hours=24)
+    cut_7d = now - timedelta(days=7)
+    cut_30d = now - timedelta(days=30)
+
+    aware_results = []
+    for r in results:
+        # mutate view via checked_at comparison with aware times
+        checked = _aware(r.checked_at)
+        aware_results.append((checked, r))
+
+    r24 = [r for checked, r in aware_results if checked >= cut_24h]
+    r7 = [r for checked, r in aware_results if checked >= cut_7d]
+    r30 = [r for checked, r in aware_results if checked >= cut_30d]
+
+    latencies = sorted(r.latency_ms for r in r7 if r.latency_ms is not None and r.response_valid)
+    valid_rate = None
+    if r7:
+        valid_rate = sum(1 for r in r7 if r.response_valid) / len(r7)
+
+    count = len(results)
+    provisional = count < 10
+    score = compute_score(
+        uptime_7d=uptime_ratio(r7),
+        latency_median_7d=float(median(latencies)) if latencies else None,
+        valid_rate_7d=valid_rate,
+        measurement_count=count,
+    )
+
+    tracking_days = (now - first_seen).total_seconds() / 86400.0
+    return TrackerMetrics(
+        uptime_24h=uptime_ratio(r24),
+        uptime_7d=uptime_ratio(r7),
+        uptime_30d=uptime_ratio(r30),
+        latency_median_7d=float(median(latencies)) if latencies else None,
+        latency_p95_7d=_percentile(list(latencies), 0.95),
+        valid_rate_7d=valid_rate,
+        measurement_count=count,
+        measurement_count_7d=len(r7),
+        tracking_days=tracking_days,
+        score=score,
+        provisional=provisional,
+    )
+
+
+async def load_metrics(session: AsyncSession, tracker: Tracker) -> TrackerMetrics:
+    result = await session.execute(
+        select(ProbeResult)
+        .where(ProbeResult.tracker_id == tracker.id)
+        .order_by(ProbeResult.checked_at.desc())
+        .limit(5000)
+    )
+    rows = list(result.scalars().all())
+    return metrics_from_results(tracker, rows)

@@ -0,0 +1,137 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+
+from sqlalchemy import select
+from sqlalchemy.ext.asyncio import AsyncSession
+from sqlalchemy.orm import selectinload
+
+from app.db.models import ProbeResult, Tracker
+from app.services.scoring import TrackerMetrics, load_metrics
+
+
+@dataclass
+class SmartListFilters:
+    protocol: str | None = None  # udp|http|https
+    min_uptime: float = 0.95
+    max_latency_ms: float | None = None
+    ip_version: str = "any"  # any|ipv4|ipv6
+    limit: int = 20
+    min_age_days: float = 0.0
+    diversity: bool = True
+    require_valid: bool = True
+    min_measurements: int = 10
+
+
+@dataclass
+class SmartListEntry:
+    tracker: Tracker
+    metrics: TrackerMetrics
+
+
+async def build_smart_list(
+    session: AsyncSession, filters: SmartListFilters | None = None
+) -> list[SmartListEntry]:
+    filters = filters or SmartListFilters()
+    limit = max(1, min(50, filters.limit))
+
+    stmt = (
+        select(Tracker).where(Tracker.current_status == "up").options(selectinload(Tracker.sources))
+    )
+    if filters.protocol:
+        stmt = stmt.where(Tracker.scheme == filters.protocol.lower())
+
+    result = await session.execute(stmt)
+    trackers = list(result.scalars().unique().all())
+
+    scored: list[SmartListEntry] = []
+    for tracker in trackers:
+        metrics = await load_metrics(session, tracker)
+        if metrics.measurement_count < filters.min_measurements:
+            continue
+        if filters.require_valid and (metrics.valid_rate_7d or 0) <= 0:
+            # require at least some valid responses; last probe should be valid
+            last = await session.execute(
+                select(ProbeResult)
+                .where(ProbeResult.tracker_id == tracker.id)
+                .order_by(ProbeResult.checked_at.desc())
+                .limit(1)
+            )
+            last_row = last.scalar_one_or_none()
+            if last_row is None or not last_row.response_valid:
+                continue
+        if metrics.uptime_7d is None or metrics.uptime_7d < filters.min_uptime:
+            continue
+        if filters.max_latency_ms is not None and (
+            metrics.latency_median_7d is None or metrics.latency_median_7d > filters.max_latency_ms
+        ):
+            continue
+        if metrics.tracking_days < filters.min_age_days:
+            continue
+        if filters.ip_version == "ipv4" and not tracker.supports_ipv4:
+            continue
+        if filters.ip_version == "ipv6" and tracker.supports_ipv6 != "true":
+            continue
+        scored.append(SmartListEntry(tracker=tracker, metrics=metrics))
+
+    scored.sort(
+        key=lambda e: (
+            -(e.metrics.score or 0),
+            e.metrics.latency_median_7d if e.metrics.latency_median_7d is not None else 1e9,
+        )
+    )
+
+    if not filters.diversity:
+        return scored[:limit]
+
+    selected: list[SmartListEntry] = []
+    seen_fps: set[str] = set()
+    asn_counts: dict[int, int] = {}
+    for entry in scored:
+        fp = entry.tracker.infrastructure_fingerprint
+        if fp and fp in seen_fps:
+            continue
+        asn = entry.tracker.asn
+        if asn is not None and asn_counts.get(asn, 0) >= 2:
+            continue
+        selected.append(entry)
+        if fp:
+            seen_fps.add(fp)
+        if asn is not None:
+            asn_counts[asn] = asn_counts.get(asn, 0) + 1
+        if len(selected) >= limit:
+            break
+    return selected
+
+
+def format_smart_list_txt(entries: list[SmartListEntry]) -> str:
+    if not entries:
+        return ""
+    return "\n\n".join(e.tracker.canonical_url for e in entries) + "\n"
+
+
+def format_smart_list_json(entries: list[SmartListEntry]) -> list[dict]:
+    out = []
+    for e in entries:
+        out.append(
+            {
+                "id": e.tracker.id,
+                "url": e.tracker.canonical_url,
+                "scheme": e.tracker.scheme,
+                "score": e.metrics.score,
+                "uptime_7d": e.metrics.uptime_7d,
+                "latency_median_7d": e.metrics.latency_median_7d,
+                "infrastructure_fingerprint": e.tracker.infrastructure_fingerprint,
+                "asn": e.tracker.asn,
+            }
+        )
+    return out
+
+
+async def build_live_list(session: AsyncSession) -> list[str]:
+    result = await session.execute(
+        select(Tracker)
+        .where(Tracker.current_status == "up")
+        .order_by(Tracker.last_latency_ms.asc().nullslast())
+    )
+    return [t.canonical_url for t in result.scalars().all()]

@@ -0,0 +1,78 @@
+from __future__ import annotations
+
+import ipaddress
+from ipaddress import IPv4Address, IPv6Address
+
+# AWS/GCP/Azure-style metadata endpoints commonly abused for SSRF
+METADATA_HOSTNAMES = frozenset(
+    {
+        "metadata.google.internal",
+        "metadata.goog",
+        "kubernetes.default",
+        "kubernetes.default.svc",
+    }
+)
+
+
+def is_blocked_hostname(hostname: str) -> bool:
+    host = hostname.strip(".").lower()
+    if host in METADATA_HOSTNAMES:
+        return True
+    if host.endswith(".local") or host.endswith(".localhost"):
+        return True
+    return host in {"localhost", "metadata"}
+
+
+def is_public_ip(ip_str: str) -> bool:
+    """Return True if the IP is globally routable / safe to connect to."""
+    try:
+        ip = ipaddress.ip_address(ip_str)
+    except ValueError:
+        return False
+    return not _is_blocked_ip(ip)
+
+
+def _is_blocked_ip(ip: IPv4Address | IPv6Address) -> bool:
+    if ip.is_loopback:
+        return True
+    if ip.is_private:
+        return True
+    if ip.is_link_local:
+        return True
+    if ip.is_multicast:
+        return True
+    if ip.is_reserved:
+        return True
+    if ip.is_unspecified:
+        return True
+    # CGNAT / documentation / benchmarking often flagged via is_private or reserved;
+    # also block classic metadata IPv4 explicitly.
+    if isinstance(ip, IPv4Address):
+        if ip == IPv4Address("169.254.169.254"):
+            return True
+        if ip in ipaddress.ip_network("0.0.0.0/8"):
+            return True
+    if isinstance(ip, IPv6Address):
+        # Unique local addresses
+        if ip.ipv4_mapped is not None:
+            return _is_blocked_ip(ip.ipv4_mapped)
+        if (int(ip) >> 118) == 0b1111110:  # fc00::/7
+            return True
+    return False
+
+
+def assert_safe_destination(hostname: str, resolved_ips: list[str]) -> None:
+    """Raise ValueError if hostname or any resolved IP is unsafe."""
+    if is_blocked_hostname(hostname):
+        raise ValueError(f"blocked hostname: {hostname}")
+    if not resolved_ips:
+        raise ValueError("no resolved addresses")
+    for ip in resolved_ips:
+        if not is_public_ip(ip):
+            raise ValueError(f"blocked address: {ip}")
+
+
+def assert_connected_ip_safe(ip_str: str) -> None:
+    """Re-check the address actually used for the connection (anti-rebinding)."""
+    if not is_public_ip(ip_str):
+        raise ValueError(f"blocked connected address: {ip_str}")

@@ -0,0 +1,346 @@
+/* Modernist-inspired theme (local recreation). No CDN. */
+
+:root {
+  --bg-outer-top: #6c7989;
+  --bg-outer-bottom: #434b55;
+  --text: #555555;
+  --text-strong: #222222;
+  --muted: #61778b;
+  --link: #006699;
+  --header-top: #ddfbfc;
+  --header-bottom: #c6eafa;
+  --header-border: #b2d2e1;
+  --btn-top: #77b9fb;
+  --btn-bottom: #3782cd;
+  --btn-border: #3a7cbe;
+  --btn-shine: #8bbef3;
+  --section-top: #fafafa;
+  --section-bottom: #dedede;
+  --wrapper-bg: #dedede;
+  --up: #2a7a2a;
+  --down: #a33;
+  --degraded: #a67c00;
+  --unknown: #666;
+  --max-width: 1220px;
+}
+
+* { box-sizing: border-box; }
+
+html {
+  background: var(--bg-outer-top);
+  background: linear-gradient(var(--bg-outer-top), var(--bg-outer-bottom)) fixed;
+  min-height: 100%;
+}
+
+body {
+  margin: 0;
+  padding: 50px 0;
+  font: 14px/1.5 Lato, "Helvetica Neue", Helvetica, Arial, sans-serif;
+  color: var(--text);
+  font-weight: 300;
+  min-height: calc(100% - 100px);
+}
+
+a {
+  color: var(--link);
+  text-decoration: none;
+}
+a:hover { text-decoration: underline; }
+a:focus-visible,
+button:focus-visible,
+input:focus-visible,
+select:focus-visible {
+  outline: 3px solid #ffbf47;
+  outline-offset: 2px;
+}
+
+.brand { color: var(--link); }
+.brand:hover { text-decoration: none; }
+
+.wrapper {
+  width: min(100% - 24px, var(--max-width));
+  margin: 0 auto;
+  background: var(--wrapper-bg);
+  border-radius: 8px;
+  box-shadow: rgba(0, 0, 0, 0.2) 0 0 0 1px, rgba(0, 0, 0, 0.45) 0 3px 10px;
+}
+
+header {
+  border-radius: 8px 8px 0 0;
+  background: linear-gradient(var(--header-top), var(--header-bottom));
+  position: relative;
+  padding: 15px 20px;
+  border-bottom: 1px solid var(--header-border);
+  min-height: 78px;
+}
+
+header h1 {
+  margin: 0;
+  padding: 0;
+  font-size: 24px;
+  line-height: 1.2;
+  color: var(--link);
+  text-shadow: rgba(255, 255, 255, 0.9) 0 1px 0;
+}
+
+header > p {
+  margin: 0;
+  color: var(--muted);
+  max-width: 42ch;
+  font-size: 13px;
+}
+
+.lang-switch {
+  margin: 6px 0 0;
+  font-size: 12px;
+  color: var(--muted);
+}
+.lang-switch a { color: var(--muted); }
+.lang-switch strong { color: var(--text-strong); }
+
+.nav-seg {
+  margin: 0;
+  padding: 1px 0;
+  list-style: none;
+  position: absolute;
+  z-index: 1;
+  right: 20px;
+  top: 20px;
+  height: 38px;
+  background: linear-gradient(var(--btn-top), var(--btn-bottom));
+  border-radius: 5px;
+  box-shadow: inset rgba(255, 255, 255, 0.45) 0 1px 0, inset rgba(0, 0, 0, 0.2) 0 -1px 0;
+  display: flex;
+}
+
+.nav-seg li {
+  border-right: 1px solid var(--btn-border);
+}
+.nav-seg li:last-child { border-right: none; }
+.nav-seg li + li { border-left: 1px solid var(--btn-shine); }
+
+.nav-seg a {
+  line-height: 1;
+  font-size: 11px;
+  color: rgba(255, 255, 255, 0.85);
+  display: block;
+  text-align: center;
+  font-weight: 400;
+  padding: 6px 12px 0;
+  height: 38px;
+  text-shadow: rgba(0, 0, 0, 0.4) 0 -1px 0;
+  text-decoration: none;
+  min-width: 72px;
+}
+.nav-seg a strong {
+  font-size: 14px;
+  display: block;
+  color: #fff;
+  font-weight: 700;
+}
+.nav-seg li.active a,
+.nav-seg a:hover {
+  background: rgba(0, 0, 0, 0.08);
+}
+
+section {
+  padding: 15px 20px 24px;
+  font-size: 15px;
+  border-top: 1px solid #fff;
+  background: linear-gradient(var(--section-top), var(--section-bottom) 700px);
+  border-radius: 0 0 8px 8px;
+}
+
+h2, h3 { color: #393939; }
+h2 { margin: 0 0 12px; }
+h3 { margin: 20px 0 8px; }
+
+.mono, code, pre, .list-out {
+  font-family: Monaco, "Bitstream Vera Sans Mono", "Lucida Console", Terminal, monospace;
+}
+
+.muted { color: var(--muted); font-size: 0.9em; }
+
+.stats-strip {
+  display: flex;
+  flex-wrap: wrap;
+  gap: 8px 16px;
+  margin: 0 0 16px;
+  padding: 8px 0;
+  border-bottom: 1px solid #aaa;
+  font-size: 13px;
+}
+.stats-strip strong { color: var(--text-strong); }
+
+.filters {
+  display: flex;
+  flex-wrap: wrap;
+  gap: 10px 14px;
+  align-items: end;
+  margin: 0 0 16px;
+}
+.filters label {
+  display: flex;
+  flex-direction: column;
+  gap: 4px;
+  font-size: 12px;
+  color: var(--muted);
+}
+.filters input,
+.filters select {
+  font: inherit;
+  padding: 4px 6px;
+  border: 1px solid #aaa;
+  border-radius: 3px;
+  background: #fff;
+  color: var(--text-strong);
+  min-width: 0;
+}
+.filters button,
+button.copy-btn {
+  font: inherit;
+  padding: 6px 12px;
+  border: 1px solid var(--btn-border);
+  border-radius: 4px;
+  color: #fff;
+  background: linear-gradient(var(--btn-top), var(--btn-bottom));
+  cursor: pointer;
+  text-shadow: rgba(0, 0, 0, 0.35) 0 -1px 0;
+}
+button.linkish {
+  background: none;
+  border: none;
+  color: var(--link);
+  cursor: pointer;
+  padding: 0 0 0 6px;
+  font: inherit;
+  font-size: 12px;
+  text-shadow: none;
+}
+
+.table-wrap {
+  width: 100%;
+  overflow-x: auto;
+  -webkit-overflow-scrolling: touch;
+}
+
+table {
+  width: 100%;
+  border-collapse: collapse;
+  font-size: 13px;
+}
+th, td {
+  text-align: left;
+  padding: 6px 8px;
+  border-bottom: 1px solid #aaa;
+  vertical-align: top;
+}
+th {
+  color: var(--text-strong);
+  font-weight: 700;
+  background: rgba(255, 255, 255, 0.45);
+}
+
+.status { white-space: nowrap; }
+.status .dot {
+  display: inline-block;
+  width: 8px;
+  height: 8px;
+  border-radius: 50%;
+  margin-right: 6px;
+  background: var(--unknown);
+  vertical-align: middle;
+}
+.status-up .dot { background: var(--up); }
+.status-down .dot { background: var(--down); }
+.status-degraded .dot { background: var(--degraded); }
+.status-unsupported .dot { background: #888; }
+
+.badge {
+  display: inline-block;
+  margin-left: 6px;
+  padding: 1px 6px;
+  border: 1px solid #aaa;
+  border-radius: 3px;
+  font-size: 11px;
+  color: var(--muted);
+}
+
+.notice {
+  background: rgba(0, 0, 0, 0.06);
+  padding: 10px 12px;
+  margin: 0 0 14px;
+}
+
+.metrics-grid {
+  display: grid;
+  grid-template-columns: repeat(auto-fill, minmax(160px, 1fr));
+  gap: 8px 12px;
+  margin: 0 0 16px;
+}
+
+.sparkline {
+  width: 100%;
+  max-width: 480px;
+  height: 48px;
+  margin: 0 0 16px;
+}
+.sparkline circle { fill: var(--link); }
+.sparkline line { stroke: var(--link); stroke-width: 1.5; }
+
+.list-out, pre {
+  background: #3a3c42;
+  color: #f8f8f2;
+  padding: 16px;
+  overflow-x: auto;
+  white-space: pre-wrap;
+  word-break: break-all;
+  font-size: 12px;
+}
+
+.pagination {
+  display: flex;
+  gap: 16px;
+  align-items: center;
+  margin-top: 12px;
+}
+
+footer {
+  width: min(100% - 24px, var(--max-width));
+  margin: 0 auto;
+  padding: 20px 0 0;
+  color: #ccc;
+  overflow: hidden;
+  display: flex;
+  justify-content: space-between;
+  gap: 12px;
+  flex-wrap: wrap;
+  font-size: 13px;
+}
+footer a { color: #fff; font-weight: 700; }
+
+@media (max-width: 900px) {
+  body { padding: 0; }
+  .wrapper {
+    width: 100%;
+    border-radius: 0;
+    box-shadow: none;
+  }
+  header { border-radius: 0; padding-bottom: 60px; }
+  .nav-seg {
+    position: static;
+    margin-top: 12px;
+    flex-wrap: wrap;
+    height: auto;
+  }
+  .nav-seg a { height: auto; padding: 8px 10px; }
+  section { border-radius: 0; }
+  footer { width: auto; padding: 16px; }
+}
+
+@media (prefers-reduced-motion: reduce) {
+  * {
+    animation: none !important;
+    transition: none !important;
+  }
+}

@@ -0,0 +1,39 @@
+(function () {
+  "use strict";
+
+  console.log("hey stranger, nothing here. kisses, pablo");
+
+  function copyText(text) {
+    if (navigator.clipboard && navigator.clipboard.writeText) {
+      return navigator.clipboard.writeText(text);
+    }
+    var ta = document.createElement("textarea");
+    ta.value = text;
+    document.body.appendChild(ta);
+    ta.select();
+    try {
+      document.execCommand("copy");
+    } finally {
+      document.body.removeChild(ta);
+    }
+    return Promise.resolve();
+  }
+
+  document.addEventListener("click", function (ev) {
+    var btn = ev.target.closest("[data-copy], [data-copy-target]");
+    if (!btn) return;
+    var text = btn.getAttribute("data-copy");
+    if (!text) {
+      var sel = btn.getAttribute("data-copy-target");
+      var el = sel ? document.querySelector(sel) : null;
+      text = el ? el.textContent : "";
+    }
+    copyText(text || "").then(function () {
+      var prev = btn.textContent;
+      btn.textContent = document.documentElement.getAttribute("data-copied") || "copied";
+      setTimeout(function () {
+        btn.textContent = prev;
+      }, 1200);
+    });
+  });
+})();

@@ -0,0 +1,44 @@
+<!DOCTYPE html>
+<html lang="{{ html_lang }}" data-copied="{{ t('copied') }}">
+<head>
+  <meta charset="utf-8">
+  <meta name="viewport" content="width=device-width, initial-scale=1">
+  <title>{% block title %}{{ app_name }}{% endblock %}</title>
+  <meta name="description" content="{{ app_description }}">
+  <link rel="stylesheet" href="/static/css/modernist.css">
+</head>
+<body>
+  <div class="wrapper">
+    <header>
+      <h1><a href="/" class="brand">{{ app_name }}</a></h1>
+      <p>{{ app_description }}</p>
+      <p class="lang-switch">
+        {% if lang == 'pt' %}<strong>pt</strong>{% else %}<a href="/lang/pt">pt</a>{% endif %}
+        ·
+        {% if lang == 'en' %}<strong>en</strong>{% else %}<a href="/lang/en">en</a>{% endif %}
+      </p>
+      <ul class="nav-seg">
+        <li class="{% if active_nav == 'trackers' %}active{% endif %}">
+          <a href="/"><strong>{{ t('nav_trackers') }}</strong>{{ t('nav_trackers_sub') }}</a>
+        </li>
+        <li class="{% if active_nav == 'lists' %}active{% endif %}">
+          <a href="/lists"><strong>{{ t('nav_lists') }}</strong>{{ t('nav_lists_sub') }}</a>
+        </li>
+        {% if repository_url %}
+        <li>
+          <a href="{{ repository_url }}" rel="noopener noreferrer"><strong>{{ t('nav_github') }}</strong>{{ t('nav_github_sub') }}</a>
+        </li>
+        {% endif %}
+      </ul>
+    </header>
+    <section>
+      {% block content %}{% endblock %}
+    </section>
+  </div>
+  <footer>
+    <p>{{ app_name }} {{ app_version }} · {{ t('footer_utc') }}</p>
+    <p>{{ t('footer_disclaimer') }}</p>
+  </footer>
+  <script src="/static/js/app.js" defer></script>
+</body>
+</html>

@@ -0,0 +1,123 @@
+{% extends "base.html" %}
+{% block title %}{{ app_name }} — {{ t('nav_trackers') }}{% endblock %}
+{% block content %}
+{% if bootstrap_note %}
+<p class="notice" role="status">{{ bootstrap_note }}</p>
+{% endif %}
+
+<div class="stats-strip" aria-label="{{ t('stats_label') }}">
+  <span><strong>{{ stats.total }}</strong> {{ t('stats_monitored') }}</span>
+  <span><strong>{{ stats.up }}</strong> {{ t('stats_up') }}</span>
+  <span><strong>{{ stats.degraded }}</strong> {{ t('stats_degraded') }}</span>
+  <span><strong>{{ stats.down }}</strong> {{ t('stats_down') }}</span>
+  <span><strong>{{ stats.unknown }}</strong> {{ t('stats_unknown') }}</span>
+  <span>{{ t('stats_last_import') }}: {{ stats.last_import_at|dt }}</span>
+  <span>{{ t('stats_last_probe') }}: {{ stats.last_probe_at|dt }}</span>
+</div>
+
+<form class="filters" method="get" action="/">
+  <label>{{ t('filter_search') }} <input type="search" name="q" value="{{ q }}" placeholder="{{ t('filter_search_ph') }}"></label>
+  <label>{{ t('filter_protocol') }}
+    <select name="scheme">
+      <option value="">{{ t('filter_any') }}</option>
+      <option value="udp" {% if scheme=='udp' %}selected{% endif %}>UDP</option>
+      <option value="http" {% if scheme=='http' %}selected{% endif %}>HTTP</option>
+      <option value="https" {% if scheme=='https' %}selected{% endif %}>HTTPS</option>
+    </select>
+  </label>
+  <label>{{ t('filter_status') }}
+    <select name="status">
+      <option value="">{{ t('filter_any') }}</option>
+      {% for s in ['up','down','degraded','unknown','unsupported'] %}
+      <option value="{{ s }}" {% if status==s %}selected{% endif %}>{{ s }}</option>
+      {% endfor %}
+    </select>
+  </label>
+  <label>{{ t('filter_source') }}
+    <select name="source">
+      <option value="">{{ t('filter_any') }}</option>
+      <option value="trackerslist" {% if source=='trackerslist' %}selected{% endif %}>trackerslist</option>
+      <option value="newtrackon" {% if source=='newtrackon' %}selected{% endif %}>newTrackon</option>
+      <option value="jokepool710" {% if source=='jokepool710' %}selected{% endif %}>jokepool710</option>
+      <option value="xiu2" {% if source=='xiu2' %}selected{% endif %}>xiu2</option>
+    </select>
+  </label>
+  <label>{{ t('filter_sort') }}
+    <select name="sort">
+      <option value="checked" {% if sort=='checked' %}selected{% endif %}>{{ t('sort_checked') }}</option>
+      <option value="latency" {% if sort=='latency' %}selected{% endif %}>{{ t('sort_latency') }}</option>
+      <option value="url" {% if sort=='url' %}selected{% endif %}>{{ t('sort_url') }}</option>
+      <option value="status" {% if sort=='status' %}selected{% endif %}>{{ t('sort_status') }}</option>
+      <option value="protocol" {% if sort=='protocol' %}selected{% endif %}>{{ t('sort_protocol') }}</option>
+    </select>
+  </label>
+  <label>{{ t('filter_order') }}
+    <select name="order">
+      <option value="desc" {% if order=='desc' %}selected{% endif %}>{{ t('order_desc') }}</option>
+      <option value="asc" {% if order=='asc' %}selected{% endif %}>{{ t('order_asc') }}</option>
+    </select>
+  </label>
+  <button type="submit">{{ t('filter_submit') }}</button>
+</form>
+
+<div class="table-wrap">
+  <table>
+    <thead>
+      <tr>
+        <th>{{ t('col_tracker') }}</th>
+        <th>{{ t('col_protocol') }}</th>
+        <th>{{ t('col_status') }}</th>
+        <th>{{ t('col_uptime7') }}</th>
+        <th>{{ t('col_latency') }}</th>
+        <th>{{ t('col_ip') }}</th>
+        <th>{{ t('col_asn') }}</th>
+        <th>{{ t('col_sources') }}</th>
+        <th>{{ t('col_aliases') }}</th>
+        <th>{{ t('col_last_check') }}</th>
+      </tr>
+    </thead>
+    <tbody>
+      {% for row in rows %}
+      <tr>
+        <td>
+          <a href="/trackers/{{ row.tracker.id }}" class="mono">{{ row.tracker.canonical_url }}</a>
+          <button type="button" class="linkish copy-btn" data-copy="{{ row.tracker.canonical_url }}">{{ t('copy') }}</button>
+          {% if row.metrics.provisional %}<span class="badge">{{ t('provisional') }}</span>{% endif %}
+        </td>
+        <td>{{ row.tracker.scheme|upper }}</td>
+        <td>
+          <span class="status status-{{ row.tracker.current_status }}" title="{{ row.tracker.current_status }}">
+            <span class="dot" aria-hidden="true"></span>{{ row.tracker.current_status }}
+          </span>
+        </td>
+        <td>{{ row.metrics.uptime_7d|pct }}</td>
+        <td>{{ row.metrics.latency_median_7d|ms }}</td>
+        <td>
+          {% if row.tracker.supports_ipv4 %}v4{% else %}—{% endif %}
+          /
+          {% if row.tracker.supports_ipv6 == 'true' %}v6{% elif row.tracker.supports_ipv6 == 'not_tested' %}v6?{% else %}—{% endif %}
+        </td>
+        <td>{% if row.tracker.asn %}AS{{ row.tracker.asn }}{% else %}—{% endif %}{% if row.tracker.network_name %} {{ row.tracker.network_name }}{% endif %}</td>
+        <td>{{ row.sources or '—' }}</td>
+        <td>{{ row.alias_count }}</td>
+        <td>{{ row.tracker.last_checked_at|dt }}</td>
+      </tr>
+      {% else %}
+      <tr><td colspan="10">{{ t('empty_inventory') }}</td></tr>
+      {% endfor %}
+    </tbody>
+  </table>
+</div>
+
+{% if pages > 1 %}
+<nav class="pagination" aria-label="{{ t('pagination') }}">
+  {% if page > 1 %}
+  <a href="?q={{ q }}&amp;scheme={{ scheme }}&amp;status={{ status }}&amp;source={{ source }}&amp;sort={{ sort }}&amp;order={{ order }}&amp;page={{ page-1 }}">{{ t('prev') }}</a>
+  {% endif %}
+  <span>{{ t('page_of', page=page, pages=pages, total=total) }}</span>
+  {% if page < pages %}
+  <a href="?q={{ q }}&amp;scheme={{ scheme }}&amp;status={{ status }}&amp;source={{ source }}&amp;sort={{ sort }}&amp;order={{ order }}&amp;page={{ page+1 }}">{{ t('next') }}</a>
+  {% endif %}
+</nav>
+{% endif %}
+{% endblock %}

@@ -0,0 +1,52 @@
+{% extends "base.html" %}
+{% block title %}{{ t('lists_title') }} — {{ app_name }}{% endblock %}
+{% block content %}
+<h2>{{ t('lists_heading') }}</h2>
+<p>{{ t('lists_blurb') }}</p>
+
+<form class="filters" method="get" action="/lists">
+  <label>{{ t('filter_protocol') }}
+    <select name="protocol">
+      <option value="">{{ t('filter_any') }}</option>
+      <option value="udp" {% if protocol=='udp' %}selected{% endif %}>UDP</option>
+      <option value="http" {% if protocol=='http' %}selected{% endif %}>HTTP</option>
+      <option value="https" {% if protocol=='https' %}selected{% endif %}>HTTPS</option>
+    </select>
+  </label>
+  <label>{{ t('min_uptime') }}
+    <input type="number" name="min_uptime" step="0.01" min="0" max="1" value="{{ min_uptime }}">
+  </label>
+  <label>{{ t('max_latency') }}
+    <input type="number" name="max_latency" min="1" value="{{ max_latency }}">
+  </label>
+  <label>{{ t('filter_ip') }}
+    <select name="ip">
+      <option value="any" {% if ip=='any' %}selected{% endif %}>{{ t('filter_any') }}</option>
+      <option value="ipv4" {% if ip=='ipv4' %}selected{% endif %}>IPv4</option>
+      <option value="ipv6" {% if ip=='ipv6' %}selected{% endif %}>IPv6</option>
+    </select>
+  </label>
+  <label>{{ t('filter_limit') }}
+    <input type="number" name="limit" min="1" max="50" value="{{ limit }}">
+  </label>
+  <label>{{ t('min_age') }}
+    <input type="number" name="min_age_days" min="0" step="0.1" value="{{ min_age_days }}">
+  </label>
+  <label>{{ t('diversity') }}
+    <select name="diversity">
+      <option value="true" {% if diversity %}selected{% endif %}>{{ t('diversity_on') }}</option>
+      <option value="false" {% if not diversity %}selected{% endif %}>{{ t('diversity_off') }}</option>
+    </select>
+  </label>
+  <button type="submit">{{ t('generate') }}</button>
+</form>
+
+<p><strong>{{ entries|length }}</strong> {{ t('selected_count') }}</p>
+<p>
+  <button type="button" class="copy-btn" data-copy-target="#list-out">{{ t('copy_btn') }}</button>
+  <a href="{{ api_txt }}">{{ t('download_txt') }}</a> ·
+  <a href="{{ api_json }}">{{ t('open_json') }}</a>
+</p>
+<p class="muted mono">{{ t('permalink') }}: {{ api_txt }}</p>
+<pre id="list-out" class="list-out">{{ txt }}</pre>
+{% endblock %}

@@ -0,0 +1,101 @@
+{% extends "base.html" %}
+{% block title %}{{ tracker.canonical_url }} — {{ app_name }}{% endblock %}
+{% block content %}
+<p><a href="/">← {{ t('back') }}</a></p>
+<h2 class="mono">{{ tracker.canonical_url }}</h2>
+<p>
+  <span class="status status-{{ tracker.current_status }}">
+    <span class="dot" aria-hidden="true"></span>{{ tracker.current_status }}
+  </span>
+  · {{ t('score') }}:
+  {% if metrics.score is not none %}{{ metrics.score }}{% else %}{{ t('score_insufficient') }}{% endif %}
+  {% if metrics.provisional %}<span class="badge">{{ t('provisional') }}</span>{% endif %}
+</p>
+
+<div class="metrics-grid">
+  <div><strong>{{ t('uptime_24h') }}</strong> {{ metrics.uptime_24h|pct }}</div>
+  <div><strong>{{ t('uptime_7d') }}</strong> {{ metrics.uptime_7d|pct }}</div>
+  <div><strong>{{ t('uptime_30d') }}</strong> {{ metrics.uptime_30d|pct }}</div>
+  <div><strong>{{ t('latency_med_7d') }}</strong> {{ metrics.latency_median_7d|ms }}</div>
+  <div><strong>{{ t('latency_p95') }}</strong> {{ metrics.latency_p95_7d|ms }}</div>
+  <div><strong>{{ t('valid_rate') }}</strong> {{ metrics.valid_rate_7d|pct }}</div>
+  <div><strong>{{ t('measurements') }}</strong> {{ metrics.measurement_count }}</div>
+  <div><strong>{{ t('tracking') }}</strong> {{ '%.1f'|format(metrics.tracking_days) }} {{ t('days') }}</div>
+</div>
+
+{% if spark %}
+<svg class="sparkline" viewBox="0 0 240 40" role="img" aria-label="{{ t('spark_label') }}">
+  {% set mx = spark|max if spark|max > 0 else 1 %}
+  {% for v in spark %}
+  {% set x = loop.index0 * (240 / ([spark|length - 1, 1]|max)) %}
+  {% set y = 38 - (v / mx * 34) %}
+  <circle cx="{{ '%.1f'|format(x) }}" cy="{{ '%.1f'|format(y) }}" r="2" />
+  {% if not loop.first %}
+  {% set px = (loop.index0 - 1) * (240 / ([spark|length - 1, 1]|max)) %}
+  {% set py = 38 - (spark[loop.index0 - 1] / mx * 34) %}
+  <line x1="{{ '%.1f'|format(px) }}" y1="{{ '%.1f'|format(py) }}" x2="{{ '%.1f'|format(x) }}" y2="{{ '%.1f'|format(y) }}" />
+  {% endif %}
+  {% endfor %}
+</svg>
+{% endif %}
+
+<h3>{{ t('sources') }}</h3>
+<ul>
+  {% for s in tracker.sources %}
+  <li>
+    {{ s.source_name }}
+    {% if s.is_best_at_source %}· {{ t('best_at_source') }}{% endif %}
+    {% if s.is_stable_at_source %}· {{ t('stable_at_source') }}{% endif %}
+    {% if s.is_live_at_source %}· {{ t('live_at_source') }}{% endif %}
+    <span class="muted">({{ t('seen') }} {{ s.last_seen_at|dt }})</span>
+  </li>
+  {% else %}
+  <li>{{ t('no_source') }}</li>
+  {% endfor %}
+</ul>
+
+<h3>{{ t('dns') }}</h3>
+<p>{{ t('terminal_cname') }}: <code>{{ tracker.terminal_cname or '—' }}</code></p>
+<p>{{ t('infra_fp') }}: <code>{{ tracker.infrastructure_fingerprint or '—' }}</code></p>
+<p>{{ t('asn') }}: {% if tracker.asn %}AS{{ tracker.asn }} {{ tracker.network_name or '' }}{% else %}{{ t('unknown') }}{% endif %}
+ · {{ t('country') }}: {{ tracker.country_code or t('unknown') }}</p>
+<ul>
+  {% for d in tracker.dns_records %}
+  <li><code>{{ d.record_type }}</code> {{ d.value }}</li>
+  {% else %}
+  <li>{{ t('no_dns') }}</li>
+  {% endfor %}
+</ul>
+
+{% if aliases %}
+<h3>{{ t('shared_infra') }}</h3>
+<p>{{ t('shared_infra_blurb') }}</p>
+<ul>
+  {% for a in aliases %}
+  <li><a class="mono" href="/trackers/{{ a.id }}">{{ a.canonical_url }}</a></li>
+  {% endfor %}
+</ul>
+{% endif %}
+
+<h3>{{ t('recent_history') }}</h3>
+<div class="table-wrap">
+  <table>
+    <thead>
+      <tr><th>{{ t('when') }}</th><th>{{ t('col_status') }}</th><th>{{ t('th_latency') }}</th><th>{{ t('valid') }}</th><th>{{ t('error') }}</th></tr>
+    </thead>
+    <tbody>
+      {% for p in probes %}
+      <tr>
+        <td>{{ p.checked_at|dt }}</td>
+        <td>{{ p.status }}</td>
+        <td>{{ p.latency_ms|ms }}</td>
+        <td>{{ t('yes') if p.response_valid else t('no') }}</td>
+        <td>{% if p.error_kind %}{{ p.error_kind }}{% if p.error_detail %}: {{ p.error_detail }}{% endif %}{% else %}—{% endif %}</td>
+      </tr>
+      {% else %}
+      <tr><td colspan="5">{{ t('no_probes') }}</td></tr>
+      {% endfor %}
+    </tbody>
+  </table>
+</div>
+{% endblock %}

@@ -0,0 +1,24 @@
+services:
+  rastro:
+    build: .
+    restart: unless-stopped
+    # All configuration lives in .env (single source of truth for prod).
+    env_file: .env
+    volumes:
+      - ./data:/app/data
+    ports:
+      # host 10359 (localhost only) -> container 8090; reverse proxy fronts TLS
+      - "127.0.0.1:10359:8090"
+    healthcheck:
+      test:
+        [
+          "CMD",
+          "python",
+          "-c",
+          "import urllib.request; urllib.request.urlopen('http://127.0.0.1:8090/healthz', timeout=3)",
+        ]
+      interval: 30s
+      timeout: 5s
+      retries: 3
+      start_period: 25s
+    stop_grace_period: 20s

@@ -0,0 +1,60 @@
+# Arquitetura
+
+## Visão geral
+
+Aplicação self-hosted de processo único que:
+
+1. importa trackers públicos de fontes externas;
+2. normaliza e protege destinos;
+3. sonda HTTP(S) e UDP;
+4. persiste histórico em SQLite;
+5. expõe HTML, TXT e JSON.
+
+Não é um tracker BitTorrent, não indexa torrents e não aceita magnets.
+
+## Processo único e Uvicorn
+
+O site, a API e o APScheduler rodam no **mesmo processo**. O Uvicorn deve ser
+iniciado com **um único worker**. Múltiplos workers duplicariam jobs de importação
+e probe. Essa limitação é intencional na versão `0.1.0`.
+
+Locks em memória impedem sobreposição do mesmo job dentro do processo.
+
+## Stack
+
+- Python 3.13, FastAPI, Jinja2, SQLAlchemy 2 async, Alembic, aiosqlite
+- SQLite com WAL
+- httpx, dnspython, APScheduler, orjson, pydantic-settings
+- Docker de um único serviço; reverse proxy TLS fica fora (Nginx/CloudPanel)
+
+## Persistência
+
+Identidade histórica do tracker é preservada mesmo quando a URL some de uma
+origem. A limpeza automática remove apenas `probe_results` expirados
+(`RAW_RESULT_RETENTION_DAYS`).
+
+Datas no banco: UTC. API: ISO 8601. UI: apresentação humana com fuso indicado.
+
+## Segurança de rede
+
+Endereços ingeridos são não confiáveis. Destinos loopback, privados, link-local,
+multicast, reservados e metadata são rejeitados. IPs do DNS são revalidados antes
+da conexão (mitigação de rebinding). Respostas têm limite de bytes e timeout.
+Não há endpoint público para testar URL arbitrária.
+
+## Configuração
+
+Nome, descrição, versão, URL pública e repositório vêm de variáveis de ambiente
+(`APP_NAME`, etc.). O nome do produto não deve ser espalhado como literal no
+código de domínio.
+
+## Degradação segura
+
+- Sem IPv6 no host: `ENABLE_IPV6=auto` marca IPv6 como não testado
+- Sem banco GeoIP: ASN/país ficam desconhecidos
+- Sem rede na primeira subida: site sobe vazio e registra o motivo
+
+## Extensões futuras (não implementadas)
+
+Contas, admin, submit público, probes distribuídos, PostgreSQL, filas, mapa,
+notificações, protocolos WS/I2P/Yggdrasil.

@@ -0,0 +1,68 @@
+# Fontes externas
+
+Este documento descreve as fontes consumidas em tempo de execução. Listas de
+origem **nunca** são apresentadas como trabalho original deste projeto.
+
+## 1. trackerslist
+
+- Repositório: <https://github.com/ngosang/trackerslist>
+- Licença: GPLv2 (apenas consumo de dados publicados; nenhum código GPLv2)
+- Arquivos:
+  - Principal: `https://raw.githubusercontent.com/ngosang/trackerslist/master/trackers_all.txt`
+  - Marcador “best”: `https://raw.githubusercontent.com/ngosang/trackerslist/master/trackers_best.txt`
+- Formato: uma URL por linha, frequentemente com linha em branco entre entradas
+- Papel no sistema: seed de endpoints públicos; `trackers_best.txt` só marca
+  `is_best_at_source`
+
+## 2. newTrackon
+
+- Repositório: <https://github.com/CorralPeltzer/newTrackon>
+- Licença: MIT
+- API:
+  - Principal: `https://newtrackon.com/api/all`
+  - Enrichment: `/api/stable`, `/api/live` (e rotas auxiliares se necessário)
+- Formato: texto plano, URLs separadas por linhas em branco
+- Papel: seed e marcadores `is_stable_at_source` / `is_live_at_source`
+- Classificações da origem **não substituem** medições próprias
+
+## 3. jokepool710/Torrent-_Trackers
+
+- Repositório: <https://github.com/jokepool710/Torrent-_Trackers>
+- Licença: MIT (© J0KEP00L, 2025)
+- Arquivo consumido:
+  - `https://raw.githubusercontent.com/jokepool710/Torrent-_Trackers/main/trackers_all.txt`
+- Formato: uma URL por linha
+- Papel: seed independente de endpoints públicos. Os arquivos por protocolo
+  (`trackers_udp`, `trackers_http_https`, `trackers_ws_wss`) são subconjuntos do
+  agregado e **não** são consumidos.
+
+## 4. XIU2/TrackersListCollection
+
+- Repositório: <https://github.com/XIU2/TrackersListCollection>
+- Licença: GPL-3.0 (apenas consumo de dados publicados; nenhum código GPL-3.0)
+- Arquivos:
+  - Principal: `https://raw.githubusercontent.com/XIU2/TrackersListCollection/master/all.txt`
+  - Marcador “best”: `https://raw.githubusercontent.com/XIU2/TrackersListCollection/master/best.txt`
+- Formato: URLs separadas por linhas em branco
+- Papel: seed agregado de vários upstreams; `best.txt` só marca
+  `is_best_at_source`. Por ser agregador, há sobreposição esperada com as demais
+  fontes — a deduplicação por URL canônica evita contagem dupla e a medição
+  permanece própria e independente.
+
+## 5. Fontes avaliadas e não adotadas
+
+- **weibone/trackers_list** (GPL-2.0): publica o mesmo conjunto de arquivos que a
+  ngosang e se comporta como espelho diário; não adiciona endpoints novos.
+
+## 6. Distinção obrigatória
+
+| Tipo de dado | Origem | Significado |
+|---|---|---|
+| Proveniência / best / stable / live | Fontes externas | Metadado de ingestão |
+| Status, latência, uptime, score | Medições locais | Resultado próprio |
+
+## 7. Política de falha
+
+Se uma origem estiver indisponível ou retornar conteúdo inválido/vazio, os dados
+anteriores são preservados. Uma resposta vazia **nunca** é interpretada como
+ordem para apagar o inventário.

@@ -0,0 +1,62 @@
+[project]
+name = "rastro"
+version = "0.1.0"
+description = "Observatorio publico de trackers BitTorrent"
+readme = "README.md"
+license = { text = "MIT" }
+authors = [{ name = "Pablo Murad" }]
+requires-python = ">=3.13"
+dependencies = [
+    "fastapi>=0.115.0",
+    "uvicorn[standard]>=0.32.0",
+    "jinja2>=3.1.4",
+    "sqlalchemy[asyncio]>=2.0.36",
+    "aiosqlite>=0.20.0",
+    "alembic>=1.14.0",
+    "httpx>=0.28.0",
+    "dnspython>=2.7.0",
+    "apscheduler>=3.10.4",
+    "orjson>=3.10.0",
+    "pydantic-settings>=2.6.0",
+]
+
+[project.optional-dependencies]
+dev = [
+    "pytest>=8.3.0",
+    "pytest-asyncio>=0.24.0",
+    "respx>=0.21.0",
+    "ruff>=0.8.0",
+    "geoip2>=4.8.0",
+]
+
+[project.scripts]
+rastro = "app.cli:main"
+
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[tool.hatch.build.targets.wheel]
+packages = ["app"]
+
+[tool.ruff]
+target-version = "py313"
+line-length = 100
+src = ["app", "tests"]
+
+[tool.ruff.lint]
+select = ["E", "F", "I", "UP", "B", "SIM"]
+ignore = ["B008"]
+
+[tool.ruff.lint.per-file-ignores]
+"alembic/*" = ["E501"]
+"tests/*" = ["E501"]
+
+[tool.pytest.ini_options]
+asyncio_mode = "auto"
+asyncio_default_fixture_loop_scope = "function"
+testpaths = ["tests"]
+filterwarnings = ["ignore::DeprecationWarning"]
+
+[tool.ruff.format]
+quote-style = "double"

@@ -0,0 +1,70 @@
+"""Shared pytest fixtures."""
+
+from __future__ import annotations
+
+import os
+from collections.abc import AsyncGenerator
+from pathlib import Path
+
+import pytest
+import pytest_asyncio
+from httpx import ASGITransport, AsyncClient
+from sqlalchemy.ext.asyncio import AsyncSession, async_sessionmaker, create_async_engine
+
+# Ensure test settings before app import
+os.environ.setdefault("APP_NAME", "RastroTest")
+os.environ.setdefault("RUN_SCHEDULER", "false")
+os.environ.setdefault("ENABLE_IPV6", "false")
+
+from app.config import Settings, get_settings
+from app.db import reset_engine
+from app.db.models import Base
+
+
+@pytest.fixture
+def settings(tmp_path: Path) -> Settings:
+    get_settings.cache_clear()
+    reset_engine()
+    db = tmp_path / "test.db"
+    s = Settings(
+        APP_NAME="RastroTest",
+        APP_DESCRIPTION="test",
+        APP_VERSION="0.1.0",
+        PUBLIC_URL="http://test.local",
+        DATABASE_URL=f"sqlite+aiosqlite:///{db.as_posix()}",
+        RUN_SCHEDULER=False,
+        ENABLE_IPV6="false",
+    )
+    return s
+
+
+@pytest_asyncio.fixture
+async def session(settings: Settings) -> AsyncGenerator[AsyncSession]:
+    engine = create_async_engine(settings.database_url, echo=False)
+    async with engine.begin() as conn:
+        await conn.run_sync(Base.metadata.create_all)
+    factory = async_sessionmaker(engine, expire_on_commit=False)
+    async with factory() as sess:
+        yield sess
+    await engine.dispose()
+
+
+@pytest_asyncio.fixture
+async def client(
+    settings: Settings, monkeypatch: pytest.MonkeyPatch
+) -> AsyncGenerator[AsyncClient]:
+    get_settings.cache_clear()
+    reset_engine()
+    monkeypatch.setenv("DATABASE_URL", settings.database_url)
+    monkeypatch.setenv("RUN_SCHEDULER", "false")
+    monkeypatch.setenv("APP_NAME", "RastroTest")
+    get_settings.cache_clear()
+
+    from app.main import create_app
+
+    application = create_app()
+    transport = ASGITransport(app=application)
+    async with AsyncClient(transport=transport, base_url="http://test") as ac:
+        yield ac
+    get_settings.cache_clear()
+    reset_engine()

@@ -0,0 +1,3 @@
+udp://fixture.example:80/announce
+
+http://fixture.example:80/announce

@@ -0,0 +1,131 @@
+"""API / HTML integration tests (offline)."""
+
+from __future__ import annotations
+
+from datetime import UTC, datetime
+
+import pytest
+from app.config import get_settings
+from app.db import reset_engine
+from app.db.models import Base, Tracker
+from app.main import create_app
+from httpx import ASGITransport, AsyncClient
+from sqlalchemy.ext.asyncio import async_sessionmaker, create_async_engine
+
+
+@pytest.fixture
+async def app_client(tmp_path, monkeypatch):
+    db = tmp_path / "api.db"
+    url = f"sqlite+aiosqlite:///{db.as_posix()}"
+    monkeypatch.setenv("DATABASE_URL", url)
+    monkeypatch.setenv("RUN_SCHEDULER", "false")
+    monkeypatch.setenv("APP_NAME", "RastroTest")
+    monkeypatch.setenv("ENABLE_IPV6", "false")
+    get_settings.cache_clear()
+    reset_engine()
+
+    engine = create_async_engine(url)
+    async with engine.begin() as conn:
+        await conn.run_sync(Base.metadata.create_all)
+    factory = async_sessionmaker(engine, expire_on_commit=False)
+    async with factory() as session:
+        session.add(
+            Tracker(
+                canonical_url="udp://demo.example:6969/announce",
+                scheme="udp",
+                hostname="demo.example",
+                port=6969,
+                path="/announce",
+                current_status="up",
+                first_seen_at=datetime.now(UTC),
+                last_seen_at=datetime.now(UTC),
+                last_checked_at=datetime.now(UTC),
+                last_latency_ms=42,
+                consecutive_failures=0,
+                infrastructure_fingerprint="abc",
+                supports_ipv4=True,
+                supports_ipv6="not_tested",
+            )
+        )
+        await session.commit()
+    await engine.dispose()
+
+    # Patch lifespan import to skip network bootstrap
+    application = create_app()
+    transport = ASGITransport(app=application)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        yield client
+    get_settings.cache_clear()
+    reset_engine()
+
+
+@pytest.mark.asyncio
+async def test_health_meta_and_pages(app_client: AsyncClient):
+    r = await app_client.get("/healthz")
+    assert r.status_code == 200
+    assert r.text == "ok"
+
+    meta = await app_client.get("/api/v1/meta")
+    assert meta.status_code == 200
+    body = meta.json()
+    assert body["name"] == "RastroTest"
+    assert "version" in body
+
+    home = await app_client.get("/")
+    assert home.status_code == 200
+    assert "demo.example" in home.text
+    assert "RastroTest" in home.text
+
+    trackers = await app_client.get("/api/v1/trackers")
+    assert trackers.status_code == 200
+    assert trackers.json()["total"] >= 1
+
+    detail = await app_client.get("/api/v1/trackers/1")
+    assert detail.status_code == 200
+
+    lists = await app_client.get("/lists")
+    assert lists.status_code == 200
+
+
+@pytest.mark.asyncio
+async def test_pages_follow_language(app_client: AsyncClient):
+    home_en = await app_client.get("/", headers={"Accept-Language": "en-US,en;q=0.9"})
+    assert home_en.status_code == 200
+    assert 'lang="en"' in home_en.text
+    assert "watched" in home_en.text
+
+    home_pt = await app_client.get(
+        "/",
+        headers={"Accept-Language": "en"},
+        cookies={"lang": "pt"},
+    )
+    assert 'lang="pt-BR"' in home_pt.text
+    assert "monitorados" in home_pt.text
+
+    missing_en = await app_client.get("/trackers/99999", headers={"Accept-Language": "en"})
+    assert missing_en.status_code == 404
+    assert "not found" in missing_en.text.lower()
+
+    missing_pt = await app_client.get("/trackers/99999", cookies={"lang": "pt"})
+    assert "não encontrado" in missing_pt.text
+
+
+@pytest.mark.asyncio
+async def test_lang_switch_sets_cookie(app_client: AsyncClient):
+    r = await app_client.get("/lang/en", follow_redirects=False)
+    assert r.status_code == 303
+    assert "lang=en" in r.headers.get("set-cookie", "")
+
+
+@pytest.mark.asyncio
+async def test_smart_list_etag(app_client: AsyncClient):
+    first = await app_client.get("/api/v1/lists/smart.txt")
+    assert first.status_code == 200
+    etag = first.headers.get("etag")
+    assert etag
+    second = await app_client.get("/api/v1/lists/smart.txt", headers={"If-None-Match": etag})
+    assert second.status_code == 304
+
+    live = await app_client.get("/api/v1/lists/live.txt")
+    assert live.status_code == 200
+    assert "demo.example" in live.text

@@ -0,0 +1,64 @@
+"""Bencode and UDP packet unit tests."""
+
+from __future__ import annotations
+
+import struct
+
+import pytest
+from app.probes.bencode import BencodeError, bdecode, validate_peers_then_discard
+from app.probes.udp import (
+    create_announce_request,
+    create_connect_request,
+    parse_announce_response,
+    parse_connect_response,
+)
+
+
+def test_bencode_valid_and_invalid():
+    # d8:intervali1800e5:peers6:\x01\x02\x03\x04\x00Pe
+    peers = b"\x01\x02\x03\x04\x00\x50"
+    payload = b"d8:intervali1800e5:peers6:" + peers + b"e"
+    decoded = bdecode(payload)
+    assert decoded["interval"] == 1800
+    assert isinstance(decoded["peers"], list)
+    assert validate_peers_then_discard(decoded) is True
+    assert decoded["peers"] is None
+
+    with pytest.raises(BencodeError):
+        bdecode(b"")
+    with pytest.raises(BencodeError):
+        bdecode(b"i123")  # incomplete
+    with pytest.raises(BencodeError):
+        bdecode(b"4:spam")  # not a dict root for tracker helper — actually bdecode requires dict
+        # 4:spam is a string root
+    with pytest.raises(BencodeError):
+        bdecode(b"l4:spame")
+
+
+def test_udp_connect_and_announce_roundtrip_fields():
+    req, tid = create_connect_request()
+    assert len(req) == 16
+    # forge response
+    conn_id = 0x1122334455667788
+    resp = struct.pack("!iiq", 0, tid, conn_id)
+    assert parse_connect_response(resp, tid) == conn_id
+
+    info = b"a" * 20
+    peer = b"b" * 20
+    ann, atid = create_announce_request(conn_id, info, peer, num_want=0)
+    assert len(ann) == 98
+    # announce response: action, tid, interval, leechers, seeders + fake peers
+    aresp = struct.pack("!iiiii", 1, atid, 1800, 2, 5) + b"\x00" * 6
+    interval, leechers, seeders = parse_announce_response(aresp, atid)
+    assert interval == 1800
+    assert leechers == 2
+    assert seeders == 5
+
+
+def test_udp_transaction_mismatch():
+    resp = struct.pack("!iiq", 0, 123, 99)
+    with pytest.raises(RuntimeError, match="transaction"):
+        parse_connect_response(resp, 456)
+    aresp = struct.pack("!iiiii", 1, 1, 10, 0, 0)
+    with pytest.raises(RuntimeError, match="transaction"):
+        parse_announce_response(aresp, 2)

@@ -0,0 +1,96 @@
+"""Collector parsing and failure isolation tests."""
+
+from __future__ import annotations
+
+import httpx
+import pytest
+import respx
+from app.collectors.base import parse_and_normalize
+from app.collectors.trackerslist import TrackerslistCollector
+from app.config import Settings
+from app.db.models import Tracker, TrackerSource
+from sqlalchemy import select
+
+FIXTURE_ALL = """
+udp://tracker.example:6969/announce
+
+http://tracker.example:80/announce
+
+udp://tracker.example:6969/announce
+
+ws://ignored.example/announce
+"""
+
+FIXTURE_BEST = """
+udp://tracker.example:6969/announce
+"""
+
+
+def test_parse_and_normalize_counts():
+    parsed = parse_and_normalize(FIXTURE_ALL)
+    assert parsed.received == 3
+    assert len(parsed.valid) == 2
+    assert len(parsed.unsupported) == 1
+
+
+@pytest.mark.asyncio
+@respx.mock
+async def test_trackerslist_import_and_empty_does_not_wipe(session, settings: Settings):
+    settings = Settings(
+        APP_NAME="RastroTest",
+        DATABASE_URL=settings.database_url,
+        TRACKERSLIST_ALL_URL="https://example.test/all.txt",
+        TRACKERSLIST_BEST_URL="https://example.test/best.txt",
+        RUN_SCHEDULER=False,
+    )
+    respx.get("https://example.test/all.txt").mock(
+        return_value=httpx.Response(200, text=FIXTURE_ALL)
+    )
+    respx.get("https://example.test/best.txt").mock(
+        return_value=httpx.Response(200, text=FIXTURE_BEST)
+    )
+    collector = TrackerslistCollector()
+    async with httpx.AsyncClient() as client:
+        stats = await collector.import_once(session, client, settings)
+    assert stats.valid == 2
+    assert stats.new == 2
+    assert stats.error is None
+
+    rows = (await session.execute(select(Tracker))).scalars().all()
+    assert len(rows) == 2
+    sources = (await session.execute(select(TrackerSource))).scalars().all()
+    assert any(s.is_best_at_source for s in sources)
+
+    # Failed empty import must preserve data
+    respx.get("https://example.test/all.txt").mock(return_value=httpx.Response(200, text="\n\n"))
+    async with httpx.AsyncClient() as client:
+        stats2 = await collector.import_once(session, client, settings)
+    assert stats2.error is not None
+    rows2 = (await session.execute(select(Tracker))).scalars().all()
+    assert len(rows2) == 2
+
+
+@pytest.mark.asyncio
+@respx.mock
+async def test_source_failure_preserves_previous(session, settings: Settings):
+    settings = Settings(
+        APP_NAME="RastroTest",
+        DATABASE_URL=settings.database_url,
+        TRACKERSLIST_ALL_URL="https://example.test/all.txt",
+        TRACKERSLIST_BEST_URL="https://example.test/best.txt",
+        RUN_SCHEDULER=False,
+    )
+    respx.get("https://example.test/all.txt").mock(
+        return_value=httpx.Response(200, text="udp://ok.example:80/announce\n")
+    )
+    respx.get("https://example.test/best.txt").mock(return_value=httpx.Response(200, text=""))
+    collector = TrackerslistCollector()
+    async with httpx.AsyncClient() as client:
+        await collector.import_once(session, client, settings)
+
+    respx.get("https://example.test/all.txt").mock(return_value=httpx.Response(503))
+    async with httpx.AsyncClient() as client:
+        stats = await collector.import_once(session, client, settings)
+    assert stats.error
+    count = len((await session.execute(select(Tracker))).scalars().all())
+    assert count == 1

@@ -0,0 +1,79 @@
+"""DNS CNAME chain / cycle unit tests with mocked resolver."""
+
+from __future__ import annotations
+
+from types import SimpleNamespace
+from unittest.mock import AsyncMock, patch
+
+import pytest
+from app.services.dns import compute_fingerprint, resolve_cname_chain
+
+
+class _Rdata:
+    def __init__(self, target: str):
+        self.target = target
+
+    def to_text(self) -> str:
+        return self.target
+
+
+@pytest.mark.asyncio
+async def test_cname_chain_and_cycle():
+    calls = {"n": 0}
+
+    async def fake_resolve(name, rdtype):
+        calls["n"] += 1
+        host = str(name).rstrip(".").lower() if not isinstance(name, str) else name.lower()
+        # dnspython passes string hostname in our code
+        host = name if isinstance(name, str) else str(name)
+        host = host.rstrip(".").lower()
+        if rdtype == "CNAME":
+            mapping = {
+                "a.example": "b.example",
+                "b.example": "a.example",  # cycle
+            }
+            if host in mapping:
+                return [_Rdata(mapping[host])]
+            raise Exception("no cname")
+        if rdtype == "A":
+            return []
+        if rdtype == "AAAA":
+            return []
+        if rdtype == "TXT":
+            raise Exception("no txt")
+        raise Exception("unexpected")
+
+    with patch("app.services.dns.dns.asyncresolver.Resolver") as resolver_cls:
+        inst = resolver_cls.return_value
+        inst.resolve = AsyncMock(side_effect=fake_resolve)
+        snap = await resolve_cname_chain("a.example")
+        assert snap.cycle_detected is True
+        assert "b.example" in snap.cname_chain
+
+
+@pytest.mark.asyncio
+async def test_cname_terminal_fingerprint():
+    async def fake_resolve(name, rdtype):
+        host = name if isinstance(name, str) else str(name)
+        host = host.rstrip(".").lower()
+        if rdtype == "CNAME":
+            if host == "start.example":
+                return [_Rdata("end.example.")]
+            raise Exception("nx")
+        if rdtype == "A":
+            if host == "end.example":
+                obj = SimpleNamespace(to_text=lambda: "203.0.113.10")
+                return [obj]
+            return []
+        if rdtype == "AAAA":
+            return []
+        if rdtype == "TXT":
+            raise Exception("no")
+        raise Exception("x")
+
+    with patch("app.services.dns.dns.asyncresolver.Resolver") as resolver_cls:
+        inst = resolver_cls.return_value
+        inst.resolve = AsyncMock(side_effect=fake_resolve)
+        snap = await resolve_cname_chain("start.example")
+        assert snap.terminal_cname == "end.example"
+        assert snap.fingerprint == compute_fingerprint("end.example", ["203.0.113.10"])

@@ -0,0 +1,45 @@
+"""HTTP size limit and scheduler lock behaviour."""
+
+from __future__ import annotations
+
+from unittest.mock import AsyncMock, patch
+
+import httpx
+import pytest
+import respx
+from app.probes.http import probe_http
+from app.scheduler import _locks, job_import_sources
+
+
+@pytest.mark.asyncio
+@respx.mock
+async def test_http_response_size_limit():
+    url = "http://tracker.example:80/announce"
+    big = b"x" * 2000
+
+    def _handler(request: httpx.Request) -> httpx.Response:
+        return httpx.Response(200, content=big)
+
+    respx.route(method="GET", host="tracker.example").mock(side_effect=_handler)
+    async with httpx.AsyncClient() as client:
+        outcome = await probe_http(
+            url,
+            client=client,
+            max_bytes=1000,
+            resolved_ips=["1.1.1.1"],
+        )
+    assert outcome.response_valid is False
+    assert outcome.error_kind == "too_large"
+
+
+@pytest.mark.asyncio
+async def test_scheduler_skips_overlapping_import():
+    # Hold the import lock and ensure job returns immediately without work
+    assert not _locks["import"].locked()
+    await _locks["import"].acquire()
+    try:
+        with patch("app.scheduler.import_all_sources", new_callable=AsyncMock) as mocked:
+            await job_import_sources()
+            mocked.assert_not_called()
+    finally:
+        _locks["import"].release()

@@ -0,0 +1,50 @@
+from __future__ import annotations
+
+from app.i18n import (
+    encode_bootstrap_note,
+    from_accept_language,
+    render_bootstrap_note,
+    resolve_lang,
+    translate,
+)
+
+
+class _Req:
+    def __init__(self, *, cookie: str | None = None, accept: str | None = None):
+        self.cookies = {"lang": cookie} if cookie else {}
+        self.headers = {"accept-language": accept} if accept else {}
+
+
+def test_cookie_wins_over_accept_language():
+    req = _Req(cookie="pt", accept="en-US,en;q=0.9")
+    assert resolve_lang(req) == "pt"
+
+
+def test_accept_language_en_without_cookie():
+    req = _Req(accept="en-US,en;q=0.8")
+    assert resolve_lang(req) == "en"
+
+
+def test_default_is_pt():
+    assert resolve_lang(_Req()) == "pt"
+    assert from_accept_language(None) == "pt"
+
+
+def test_bootstrap_note_follows_lang():
+    raw = encode_bootstrap_note("empty_import", "boom")
+    pt = render_bootstrap_note("pt", raw)
+    en = render_bootstrap_note("en", raw)
+    assert "Inventário vazio" in pt
+    assert "boom" in pt
+    assert "Empty inventory" in en
+    assert "boom" in en
+
+
+def test_legacy_bootstrap_note_passthrough():
+    old = "Inventário vazio: alguma coisa"
+    assert render_bootstrap_note("en", old) == old
+
+
+def test_translate_fallback():
+    assert translate("en", "not_found") == "Tracker not found"
+    assert translate("pt", "not_found") == "Tracker não encontrado"

@@ -0,0 +1,112 @@
+"""Tests for the jokepool710 and XIU2 collectors and the shared feed runner."""
+
+from __future__ import annotations
+
+import httpx
+import pytest
+import respx
+from app.collectors.jokepool710 import Jokepool710Collector
+from app.collectors.xiu2 import Xiu2Collector
+from app.config import Settings
+from app.db.models import Tracker, TrackerSource
+from sqlalchemy import select
+
+FIXTURE_JOKEPOOL = """
+udp://tracker.example:6969/announce
+http://tracker.example:80/announce
+udp://tracker.example:6969/announce
+"""
+
+FIXTURE_XIU2_ALL = """
+udp://a.example:6969/announce
+
+http://b.example:80/announce
+"""
+
+FIXTURE_XIU2_BEST = """
+udp://a.example:6969/announce
+"""
+
+
+@pytest.mark.asyncio
+@respx.mock
+async def test_jokepool710_import_and_empty_does_not_wipe(session, settings: Settings):
+    settings = Settings(
+        APP_NAME="RastroTest",
+        DATABASE_URL=settings.database_url,
+        JOKEPOOL710_ALL_URL="https://example.test/joke.txt",
+        RUN_SCHEDULER=False,
+    )
+    respx.get("https://example.test/joke.txt").mock(
+        return_value=httpx.Response(200, text=FIXTURE_JOKEPOOL)
+    )
+    collector = Jokepool710Collector()
+    async with httpx.AsyncClient() as client:
+        stats = await collector.import_once(session, client, settings)
+    assert stats.valid == 2
+    assert stats.new == 2
+    assert stats.error is None
+
+    rows = (await session.execute(select(Tracker))).scalars().all()
+    assert len(rows) == 2
+    sources = (await session.execute(select(TrackerSource))).scalars().all()
+    assert {s.source_name for s in sources} == {"jokepool710"}
+
+    # Empty response must preserve existing data.
+    respx.get("https://example.test/joke.txt").mock(return_value=httpx.Response(200, text="\n\n"))
+    async with httpx.AsyncClient() as client:
+        stats2 = await collector.import_once(session, client, settings)
+    assert stats2.error is not None
+    rows2 = (await session.execute(select(Tracker))).scalars().all()
+    assert len(rows2) == 2
+
+
+@pytest.mark.asyncio
+@respx.mock
+async def test_xiu2_import_sets_best_flag(session, settings: Settings):
+    settings = Settings(
+        APP_NAME="RastroTest",
+        DATABASE_URL=settings.database_url,
+        XIU2_ALL_URL="https://example.test/xiu2-all.txt",
+        XIU2_BEST_URL="https://example.test/xiu2-best.txt",
+        RUN_SCHEDULER=False,
+    )
+    respx.get("https://example.test/xiu2-all.txt").mock(
+        return_value=httpx.Response(200, text=FIXTURE_XIU2_ALL)
+    )
+    respx.get("https://example.test/xiu2-best.txt").mock(
+        return_value=httpx.Response(200, text=FIXTURE_XIU2_BEST)
+    )
+    collector = Xiu2Collector()
+    async with httpx.AsyncClient() as client:
+        stats = await collector.import_once(session, client, settings)
+    assert stats.valid == 2
+    assert stats.new == 2
+    assert stats.error is None
+
+    sources = (await session.execute(select(TrackerSource))).scalars().all()
+    best = [s for s in sources if s.is_best_at_source]
+    assert len(best) == 1
+
+
+@pytest.mark.asyncio
+@respx.mock
+async def test_xiu2_best_failure_does_not_fail_import(session, settings: Settings):
+    settings = Settings(
+        APP_NAME="RastroTest",
+        DATABASE_URL=settings.database_url,
+        XIU2_ALL_URL="https://example.test/xiu2-all.txt",
+        XIU2_BEST_URL="https://example.test/xiu2-best.txt",
+        RUN_SCHEDULER=False,
+    )
+    respx.get("https://example.test/xiu2-all.txt").mock(
+        return_value=httpx.Response(200, text=FIXTURE_XIU2_ALL)
+    )
+    respx.get("https://example.test/xiu2-best.txt").mock(return_value=httpx.Response(503))
+    collector = Xiu2Collector()
+    async with httpx.AsyncClient() as client:
+        stats = await collector.import_once(session, client, settings)
+    assert stats.error is None
+    assert stats.valid == 2
+    sources = (await session.execute(select(TrackerSource))).scalars().all()
+    assert all(not s.is_best_at_source for s in sources)

@@ -0,0 +1,60 @@
+"""Normalization and SSRF unit tests."""
+
+from __future__ import annotations
+
+import pytest
+from app.services.normalize import (
+    NormalizationError,
+    normalize_tracker_url,
+    parse_tracker_list_text,
+)
+from app.services.ssrf import assert_safe_destination, is_blocked_hostname, is_public_ip
+
+
+def test_parse_blank_lines_and_duplicates():
+    body = "udp://a.example:80/announce\n\n\nudp://a.example:80/announce\nhttp://b.example:80/announce\n"
+    items = parse_tracker_list_text(body)
+    assert items == [
+        "udp://a.example:80/announce",
+        "http://b.example:80/announce",
+    ]
+
+
+def test_normalize_basic_and_idna():
+    n = normalize_tracker_url("HTTPS://EXEMPLO.COM:443/announce")
+    assert n.scheme == "https"
+    assert n.hostname == "exemplo.com"
+    assert n.port == 443
+    assert n.canonical_url == "https://exemplo.com:443/announce"
+
+    n2 = normalize_tracker_url("http://bücher.example/announce")
+    assert "xn--" in n2.hostname
+    assert n2.scheme == "http"
+
+
+def test_reject_credentials_and_bad_port():
+    with pytest.raises(NormalizationError):
+        normalize_tracker_url("http://user:pass@host.example:80/announce")
+    with pytest.raises(NormalizationError):
+        normalize_tracker_url("udp://host.example:99999/announce")
+
+
+def test_unsupported_ws():
+    n = normalize_tracker_url("wss://tracker.example/announce")
+    assert n.unsupported is True
+
+
+def test_ssrf_blocks_private_and_metadata():
+    assert is_public_ip("8.8.8.8")
+    assert not is_public_ip("127.0.0.1")
+    assert not is_public_ip("10.0.0.1")
+    assert not is_public_ip("192.168.1.1")
+    assert not is_public_ip("169.254.169.254")
+    assert not is_public_ip("::1")
+    assert not is_public_ip("fc00::1")
+    assert is_blocked_hostname("localhost")
+    assert is_blocked_hostname("metadata.google.internal")
+    with pytest.raises(ValueError):
+        assert_safe_destination("evil.example", ["127.0.0.1"])
+    with pytest.raises(ValueError):
+        assert_safe_destination("localhost", ["8.8.8.8"])

@@ -0,0 +1,80 @@
+"""Weekly import cron trigger and import-staleness helper."""
+
+from __future__ import annotations
+
+from datetime import UTC, datetime, timedelta
+
+import pytest
+from app.config import Settings
+from app.db.models import SourceImportRun
+from app.db.queries import latest_successful_import_at
+from app.scheduler import start_scheduler, stop_scheduler
+
+
+@pytest.mark.asyncio
+async def test_import_job_registered_as_weekly_cron():
+    settings = Settings(
+        APP_NAME="RastroTest",
+        RUN_SCHEDULER=True,
+        IMPORT_DAY_OF_WEEK="sun",
+        IMPORT_HOUR_UTC=4,
+    )
+    scheduler = start_scheduler(settings)
+    try:
+        job = scheduler.get_job("import_sources")
+        assert job is not None
+        trigger = str(job.trigger)
+        assert "day_of_week='sun'" in trigger
+        assert "hour='4'" in trigger
+    finally:
+        stop_scheduler()
+
+
+@pytest.mark.asyncio
+async def test_latest_successful_import_at(session):
+    now = datetime.now(UTC)
+    session.add_all(
+        [
+            SourceImportRun(
+                source_name="a",
+                started_at=now - timedelta(days=10),
+                finished_at=now - timedelta(days=10),
+                status="ok",
+            ),
+            SourceImportRun(
+                source_name="b",
+                started_at=now - timedelta(days=2),
+                finished_at=now - timedelta(days=2),
+                status="not_modified",
+            ),
+            SourceImportRun(
+                source_name="c",
+                started_at=now,
+                finished_at=now,
+                status="failed",
+            ),
+        ]
+    )
+    await session.commit()
+
+    latest = await latest_successful_import_at(session)
+    assert latest is not None
+    if latest.tzinfo is None:
+        latest = latest.replace(tzinfo=UTC)
+    # The 2-day-old not_modified run wins; the failed "now" run does not count.
+    assert (now - latest) < timedelta(days=3)
+
+
+@pytest.mark.asyncio
+async def test_latest_successful_import_at_none_when_only_failures(session):
+    now = datetime.now(UTC)
+    session.add(
+        SourceImportRun(
+            source_name="a",
+            started_at=now,
+            finished_at=now,
+            status="failed",
+        )
+    )
+    await session.commit()
+    assert await latest_successful_import_at(session) is None

@@ -0,0 +1,106 @@
+"""DNS fingerprint helpers and scoring / smart list."""
+
+from __future__ import annotations
+
+from datetime import UTC, datetime, timedelta
+
+from app.db.models import ProbeResult, Tracker
+from app.services.dns import compute_fingerprint
+from app.services.scoring import compute_score, latency_score, metrics_from_results
+from app.services.smart_list import SmartListEntry, SmartListFilters, format_smart_list_txt
+
+
+def test_cname_fingerprint_deterministic():
+    a = compute_fingerprint("cdn.example.net", ["1.2.3.4"])
+    b = compute_fingerprint("cdn.example.net", ["9.9.9.9"])
+    assert a == b
+    ips = compute_fingerprint(None, ["2.2.2.2", "1.1.1.1"])
+    ips2 = compute_fingerprint(None, ["1.1.1.1", "2.2.2.2"])
+    assert ips == ips2
+    assert a != ips
+
+
+def test_scoring_and_provisional():
+    assert latency_score(100) == 20
+    assert latency_score(250) == 16
+    assert latency_score(2000) == 0
+    assert (
+        compute_score(uptime_7d=1.0, latency_median_7d=100, valid_rate_7d=1.0, measurement_count=2)
+        is None
+    )
+    score = compute_score(
+        uptime_7d=1.0, latency_median_7d=100, valid_rate_7d=1.0, measurement_count=25
+    )
+    assert score == 100
+
+    tracker = Tracker(
+        canonical_url="udp://t.example:80/announce",
+        scheme="udp",
+        hostname="t.example",
+        port=80,
+        path="/announce",
+        current_status="up",
+        first_seen_at=datetime.now(UTC) - timedelta(days=3),
+        last_seen_at=datetime.now(UTC),
+        consecutive_failures=0,
+    )
+    now = datetime.now(UTC)
+    results = [
+        ProbeResult(
+            tracker_id=1,
+            checked_at=now - timedelta(hours=i),
+            status="up",
+            latency_ms=120,
+            response_valid=True,
+        )
+        for i in range(5)
+    ]
+    m = metrics_from_results(tracker, results, now=now)
+    assert m.provisional is True
+    assert m.score is not None
+    assert m.uptime_7d == 1.0
+
+
+def test_smart_list_txt_blank_lines():
+    t1 = Tracker(
+        id=1,
+        canonical_url="udp://a.example:80/announce",
+        scheme="udp",
+        hostname="a.example",
+        port=80,
+        path="/announce",
+        current_status="up",
+        first_seen_at=datetime.now(UTC),
+        last_seen_at=datetime.now(UTC),
+        consecutive_failures=0,
+    )
+    t2 = Tracker(
+        id=2,
+        canonical_url="udp://b.example:80/announce",
+        scheme="udp",
+        hostname="b.example",
+        port=80,
+        path="/announce",
+        current_status="up",
+        first_seen_at=datetime.now(UTC),
+        last_seen_at=datetime.now(UTC),
+        consecutive_failures=0,
+    )
+    from app.services.scoring import TrackerMetrics
+
+    metrics = TrackerMetrics(
+        uptime_24h=1,
+        uptime_7d=1,
+        uptime_30d=1,
+        latency_median_7d=10,
+        latency_p95_7d=20,
+        valid_rate_7d=1,
+        measurement_count=20,
+        measurement_count_7d=20,
+        tracking_days=10,
+        score=90,
+        provisional=False,
+    )
+    txt = format_smart_list_txt([SmartListEntry(t1, metrics), SmartListEntry(t2, metrics)])
+    assert txt == "udp://a.example:80/announce\n\nudp://b.example:80/announce\n"
+    assert SmartListFilters().limit == 20

@@ -0,0 +1,71 @@
+"""Smart list diversity selection."""
+
+from __future__ import annotations
+
+from datetime import UTC, datetime, timedelta
+
+import pytest
+from app.db.models import ProbeResult, Tracker
+from app.services.smart_list import SmartListFilters, build_smart_list
+from sqlalchemy.ext.asyncio import AsyncSession
+
+
+async def _seed_tracker(
+    session: AsyncSession,
+    *,
+    url: str,
+    fp: str,
+    asn: int | None,
+    latency: float,
+) -> Tracker:
+    now = datetime.now(UTC)
+    t = Tracker(
+        canonical_url=url,
+        scheme="udp",
+        hostname=url.split("://")[1].split(":")[0],
+        port=80,
+        path="/announce",
+        current_status="up",
+        first_seen_at=now - timedelta(days=30),
+        last_seen_at=now,
+        last_checked_at=now,
+        last_latency_ms=latency,
+        consecutive_failures=0,
+        infrastructure_fingerprint=fp,
+        asn=asn,
+        supports_ipv4=True,
+        supports_ipv6="not_tested",
+    )
+    session.add(t)
+    await session.flush()
+    for i in range(12):
+        session.add(
+            ProbeResult(
+                tracker_id=t.id,
+                checked_at=now - timedelta(hours=i),
+                status="up",
+                latency_ms=latency,
+                response_valid=True,
+            )
+        )
+    await session.commit()
+    return t
+
+
+@pytest.mark.asyncio
+async def test_smart_list_diversity(session: AsyncSession):
+    await _seed_tracker(session, url="udp://a1.example:80/announce", fp="fp1", asn=100, latency=50)
+    await _seed_tracker(session, url="udp://a2.example:80/announce", fp="fp1", asn=100, latency=40)
+    await _seed_tracker(session, url="udp://b1.example:80/announce", fp="fp2", asn=100, latency=60)
+    await _seed_tracker(session, url="udp://c1.example:80/announce", fp="fp3", asn=200, latency=70)
+
+    entries = await build_smart_list(
+        session, SmartListFilters(min_uptime=0.5, limit=20, diversity=True)
+    )
+    urls = [e.tracker.canonical_url for e in entries]
+    # only one of fp1
+    assert sum(1 for u in urls if u.startswith("udp://a")) == 1
+    # asn 100 at most twice
+    asn100 = [e for e in entries if e.tracker.asn == 100]
+    assert len(asn100) <= 2
+    assert "udp://c1.example:80/announce" in urls