feat(docker): per-host collection via the host agent; drop central scrape

Docker collection moves off the central single-socket scrape onto the host
agent, giving Docker a real per-host dimension. The Steward host now reports
its own containers like any other host, and same-named containers on different
hosts no longer collide.

- agent: stdlib UDS Docker client (AF_UNIX HTTP/1.1, Connection: close,
  chunked-aware), collect_docker() ports the cpu%/mem math; sample["docker"]
  added best-effort (silent-skip on absent/unreadable socket). AGENT_VERSION
  1.2.0 → 1.3.0; optional docker_socket config key.
- ingest: host_agent ingest hands per-host container snapshots to the docker
  plugin via a new "docker.persist_host_samples" capability (no hard import,
  no-op when docker disabled), inside a SAVEPOINT so a docker failure never
  sinks the host metrics. Resource names are host-scoped ("<host>/<name>").
- schema: docker_containers re-keyed (host_id, name); docker_metrics gains
  host_id; docker_002 migration DROP+recreates (dev-only, rule 122).
- ui: Docker page + widgets grouped by host with host links; new per-host
  Docker panel embedded on the Hosts hub (gated on docker enabled via a new
  enabled_plugins template context). Replaces the SQLite-only strftime
  bucketing with DB-agnostic Python bucketing.
- provisioning: install/provision playbooks add steward-agent to the docker
  group (best-effort) so the agent can read the socket.
- removed central scrape: docker scheduler.py + scraper.py deleted; plugin.yaml
  socket_path/scrape_interval_seconds/include_stopped dropped (plugin 2.0.0).
- tests: agent docker collector units (math, chunked decode, silent-skip,
  sample shape, config) + integration (host-scoped schema + persistence).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016Jg27rgypiW2efULXJDtMC

plugins/docker/models.py

@@ -2,16 +2,24 @@
 from __future__ import annotations
 import uuid
 from datetime import datetime, timezone
-from sqlalchemy import DateTime, Float, Integer, String, Text
+from sqlalchemy import DateTime, Float, ForeignKey, Index, Integer, String, Text
 from sqlalchemy.orm import Mapped, mapped_column
 from steward.models.base import Base
 
 
 class DockerContainer(Base):
-    """Latest known state per container — upserted by name on each scrape."""
+    """Latest known state per container, scoped to the host that reported it.
+
+    Collection is per-host via the host agent, so container names are only
+    unique within a host — the natural key is (host_id, name). host_id is NOT
+    NULL: every container arrives through a host_agent ingest that resolves a
+    Host first. Deleting a host cascades its containers away.
+    """
     __tablename__ = "docker_containers"
 
-    # Container name is the stable natural key; container_id changes on recreation
+    host_id: Mapped[str] = mapped_column(
+        String(36), ForeignKey("hosts.id", ondelete="CASCADE"), primary_key=True
+    )
     name: Mapped[str] = mapped_column(String(255), primary_key=True)
     container_id: Mapped[str] = mapped_column(String(64), nullable=False, default="")
     image: Mapped[str] = mapped_column(String(512), nullable=False, default="")
@@ -32,12 +40,16 @@ class DockerContainer(Base):
 
 
 class DockerMetric(Base):
-    """Time-series CPU/memory per container — one row per scrape per running container."""
+    """Time-series CPU/memory per container — one row per sample per running
+    container, scoped to the reporting host."""
     __tablename__ = "docker_metrics"
 
     id: Mapped[str] = mapped_column(
         String(36), primary_key=True, default=lambda: str(uuid.uuid4())
     )
+    host_id: Mapped[str] = mapped_column(
+        String(36), ForeignKey("hosts.id", ondelete="CASCADE"), nullable=False, index=True
+    )
     container_name: Mapped[str] = mapped_column(String(255), nullable=False, index=True)
     scraped_at: Mapped[datetime] = mapped_column(
         DateTime(timezone=True), nullable=False,
@@ -47,3 +59,10 @@ class DockerMetric(Base):
     cpu_pct: Mapped[float] = mapped_column(Float, nullable=False, default=0.0)
     mem_pct: Mapped[float] = mapped_column(Float, nullable=False, default=0.0)
     mem_usage_bytes: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
+
+    # Per-container history lookups filter on (host_id, container_name) then sort
+    # by time — a composite index keeps the rows() sparkline queries cheap.
+    __table_args__ = (
+        Index("ix_docker_metrics_host_container_time",
+              "host_id", "container_name", "scraped_at"),
+    )