feat(knowledge): two-tier pagination — ID pre-fetch + content batch loading

Backend:
- GET /api/knowledge/ids: returns up to 100 note IDs cheaply (no body
  parsing), supports same filters as /api/knowledge, includes has_more
- GET /api/knowledge/batch?ids=...: fetches full items for given IDs in
  order; used by frontend to load content in controlled batches

Frontend (KnowledgeView):
- Fetch 100 IDs upfront, load first 50 as content on mount
- IntersectionObserver sentinel (root: null) triggers 24-item content
  batches as user scrolls
- Proactive ID refill when queue drops below 48 unloaded IDs
- fetchGen counter invalidates stale in-flight responses on filter reset
- IDs claimed before async fetch to prevent double-loading
- sentinelVisible ref drives post-load re-check when content doesn't
  push sentinel off screen

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

src/fabledassistant/routes/knowledge.py

@@ -62,6 +62,63 @@ async def list_knowledge():
     })
 
 
+@knowledge_bp.route("/ids", methods=["GET"])
+@login_required
+async def list_knowledge_ids():
+    """Return note IDs only (cheap) for the two-tier pagination feed.
+
+    Same filter params as GET /api/knowledge.
+    Additional params: limit (default 100, max 200), offset (default 0).
+    Returns {ids, total, has_more}.
+    """
+    uid = get_current_user_id()
+    note_type = request.args.get("type", "").strip().lower() or None
+    tags_raw = request.args.get("tags", "").strip()
+    tags = [t.strip() for t in tags_raw.split(",") if t.strip()] if tags_raw else []
+    sort = request.args.get("sort", "modified").strip().lower()
+    q = request.args.get("q", "").strip() or None
+    if sort not in _VALID_SORTS:
+        sort = "modified"
+    try:
+        limit = min(int(request.args.get("limit", 100)), 200)
+        offset = max(0, int(request.args.get("offset", 0)))
+    except ValueError:
+        return jsonify({"error": "Invalid limit or offset"}), 400
+
+    if note_type and note_type not in _VALID_TYPES:
+        return jsonify({"error": "Invalid type"}), 400
+
+    from fabledassistant.services.knowledge import query_knowledge_ids
+    ids, total = await query_knowledge_ids(
+        user_id=uid, note_type=note_type, tags=tags,
+        sort=sort, q=q, limit=limit, offset=offset,
+    )
+    return jsonify({"ids": ids, "total": total, "has_more": (offset + len(ids)) < total})
+
+
+@knowledge_bp.route("/batch", methods=["GET"])
+@login_required
+async def get_knowledge_batch():
+    """Fetch full items for a comma-separated list of IDs (max 100).
+
+    Returns {items: [...]} in the order of the requested IDs.
+    """
+    uid = get_current_user_id()
+    ids_raw = request.args.get("ids", "").strip()
+    if not ids_raw:
+        return jsonify({"items": []})
+    try:
+        ids = [int(x) for x in ids_raw.split(",") if x.strip()]
+    except ValueError:
+        return jsonify({"error": "Invalid IDs"}), 400
+    if len(ids) > 100:
+        return jsonify({"error": "Too many IDs (max 100)"}), 400
+
+    from fabledassistant.services.knowledge import get_knowledge_by_ids
+    items = await get_knowledge_by_ids(uid, ids)
+    return jsonify({"items": items})
+
+
 @knowledge_bp.route("/tags", methods=["GET"])
 @login_required
 async def list_knowledge_tags():

src/fabledassistant/services/knowledge.py

@@ -167,6 +167,69 @@ async def get_knowledge_tags(user_id: int, note_type: str | None = None) -> list
     return [r for r in rows if r]
 
 
+async def query_knowledge_ids(
+    user_id: int,
+    note_type: str | None,
+    tags: list[str],
+    sort: str,
+    q: str | None,
+    limit: int = 100,
+    offset: int = 0,
+) -> tuple[list[int], int]:
+    """Return note IDs only — cheap query for the two-tier pagination feed."""
+    if q:
+        # Re-use semantic search, extract IDs in rank order
+        items, total = await _semantic_knowledge_search(
+            user_id, q, note_type=note_type, tags=tags,
+            limit=limit, offset=offset,
+        )
+        return [item["id"] for item in items], total
+
+    async with async_session() as session:
+        base = (
+            select(Note.id)
+            .where(Note.user_id == user_id)
+            .where(Note.status.is_(None))
+        )
+        if note_type:
+            base = base.where(Note.note_type == note_type)
+        else:
+            base = base.where(Note.note_type.in_(["note", "person", "place", "list"]))
+        for tag in tags:
+            base = base.where(Note.tags.contains([tag]))
+
+        count_stmt = select(func.count()).select_from(base.subquery())
+        total: int = (await session.execute(count_stmt)).scalar_one()
+
+        if sort == "created":
+            base = base.order_by(Note.created_at.desc())
+        elif sort == "alpha":
+            base = base.order_by(Note.title.asc())
+        elif sort == "type":
+            base = base.order_by(Note.note_type.asc(), Note.updated_at.desc())
+        else:
+            base = base.order_by(Note.updated_at.desc())
+
+        ids = list((await session.execute(base.limit(limit).offset(offset))).scalars().all())
+
+    return ids, total
+
+
+async def get_knowledge_by_ids(user_id: int, ids: list[int]) -> list[dict]:
+    """Fetch full items for the given IDs, preserving the requested order."""
+    if not ids:
+        return []
+    async with async_session() as session:
+        stmt = (
+            select(Note)
+            .where(Note.user_id == user_id)
+            .where(Note.id.in_(ids))
+        )
+        rows = list((await session.execute(stmt)).scalars().all())
+    by_id = {n.id: n for n in rows}
+    return [_note_to_item(by_id[i]) for i in ids if i in by_id]
+
+
 async def get_people_and_places_context(user_id: int) -> str:
     """Return a compact summary of known people and places for LLM system prompt injection."""
     async with async_session() as session: