feat(plugin): knowledge auto-inject (Path A) — title-first per-turn awareness

New UserPromptSubmit hook (scribe_autoinject.sh) + GET /api/plugin/retrieve that
surface the TITLES (never bodies) of the few notes clearing four anti-bloat
gates: a per-user confidence threshold (stricter than pull search), a margin
gate, per-session dedup (exclude_ids), and a top-k ceiling. Each retrieval is
logged to retrieval_logs as source=auto_inject so the threshold can be tuned
from data. Per-user config (enable / threshold / top-k) is DB-backed via
/api/settings with a Settings UI card; defaults enabled, threshold 0.55,
top-k 3 (conservative — tune once auto_inject telemetry accrues).

Scribe: project 2, milestone 93, task 1033.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Xz4j1H7pjYSjKsEpgcNH5E

tests/test_services_plugin_context.py

@@ -10,6 +10,92 @@ def _rule(rid, title, topic_id):
     return r
 
 
+def _note(nid, title):
+    n = MagicMock()
+    n.id, n.title = nid, title
+    return n
+
+
+# ─── knowledge auto-inject (Path A) ──────────────────────────────────────────
+
+
+@pytest.mark.asyncio
+async def test_get_autoinject_config_defaults_and_clamps():
+    from scribe.services import plugin_context as pc
+
+    # No settings stored → defaults.
+    with patch.object(pc, "get_setting", AsyncMock(side_effect=lambda uid, k, d: d)):
+        cfg = await pc.get_autoinject_config(1)
+    assert cfg == {
+        "enabled": pc.AUTOINJECT_DEFAULT_ENABLED,
+        "threshold": pc.AUTOINJECT_DEFAULT_THRESHOLD,
+        "top_k": pc.AUTOINJECT_DEFAULT_TOP_K,
+    }
+
+    # Out-of-range values are clamped; top_k capped at the hard ceiling.
+    stored = {
+        pc.AUTOINJECT_ENABLED_KEY: "false",
+        pc.AUTOINJECT_THRESHOLD_KEY: "5",
+        pc.AUTOINJECT_TOP_K_KEY: "999",
+    }
+    with patch.object(pc, "get_setting",
+                      AsyncMock(side_effect=lambda uid, k, d: stored.get(k, d))):
+        cfg = await pc.get_autoinject_config(1)
+    assert cfg["enabled"] is False
+    assert cfg["threshold"] == 1.0
+    assert cfg["top_k"] == pc._AUTOINJECT_MAX_TOP_K
+
+
+@pytest.mark.asyncio
+async def test_build_autoinject_hint_disabled_returns_empty_and_skips_search():
+    from scribe.services import plugin_context as pc
+    search = AsyncMock()
+    with patch.object(pc, "get_autoinject_config",
+                      AsyncMock(return_value={"enabled": False, "threshold": 0.55, "top_k": 3})), \
+         patch.object(pc, "semantic_search_notes", search), \
+         patch.object(pc, "record_retrieval", MagicMock()):
+        out = await pc.build_autoinject_hint(1, "anything")
+    assert out["context"] == "" and out["note_ids"] == []
+    search.assert_not_called()  # disabled → no retrieval at all
+
+
+@pytest.mark.asyncio
+async def test_build_autoinject_hint_titles_only_with_margin_gate():
+    from scribe.services import plugin_context as pc
+    # top=0.80; 0.74 within band (0.10), 0.61 outside → dropped.
+    hits = [(0.80, _note(11, "Pool sizing decision")),
+            (0.74, _note(22, "run_maintenance thresholds")),
+            (0.61, _note(33, "unrelated-ish"))]
+    rec = MagicMock()
+    with patch.object(pc, "get_autoinject_config",
+                      AsyncMock(return_value={"enabled": True, "threshold": 0.55, "top_k": 3})), \
+         patch.object(pc, "semantic_search_notes", AsyncMock(return_value=hits)), \
+         patch.object(pc, "record_retrieval", rec):
+        out = await pc.build_autoinject_hint(1, "postgres pool", project_id=2,
+                                             exclude_ids=[99])
+    # Margin gate kept the top two, dropped the straggler.
+    assert out["note_ids"] == [11, 22]
+    assert '#11 "Pool sizing decision" (0.80)' in out["context"]
+    assert "#33" not in out["context"]
+    # Title-first: no body text, ever.
+    assert "get_note(id)" in out["context"]
+    # Telemetry fired with the auto_inject source and the full candidate set.
+    rec.assert_called_once()
+    assert rec.call_args.kwargs["source"] == "auto_inject"
+
+
+@pytest.mark.asyncio
+async def test_build_autoinject_hint_blank_query_returns_empty():
+    from scribe.services import plugin_context as pc
+    search = AsyncMock()
+    with patch.object(pc, "get_autoinject_config",
+                      AsyncMock(return_value={"enabled": True, "threshold": 0.55, "top_k": 3})), \
+         patch.object(pc, "semantic_search_notes", search):
+        out = await pc.build_autoinject_hint(1, "   ")
+    assert out["context"] == ""
+    search.assert_not_called()
+
+
 @pytest.mark.asyncio
 async def test_build_session_context_renders_titles_grouped_by_topic():
     rules = [