feat: add deepsearch research system with CLI prune/clear and database schema

Implement a multi-agent deep web research subsystem including CLI commands for pruning expired jobs and clearing all artifacts, database tables for sessions/messages/URL cache with indexes, config paths for chroma storage, and internal embed URL for vector operations.

tests/unit/services/openai_gateway/gateway.py

@@ -246,6 +246,36 @@ def test_embeddings_remaps_alias_to_configured_model(local_db, monkeypatch):
     assert rt._client.calls[-1][1]["model"] == "qwen/qwen3-embedding-8b"
 
 
+def test_embeddings_success_records_ledger(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeEmbedClient_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    cfg["gateway_force_model"] = True
+    cfg["gateway_embed_enabled"] = True
+    cfg["gateway_embed_model"] = "qwen/qwen3-embedding-8b"
+    rt = svc.runtime()
+    before = rt.embed_calls
+    resp = run_async(
+        rt.handle_embeddings(
+            {"model": "molodetz~embed", "input": "hello"},
+            cfg,
+            ("guest", "ledger_probe"),
+            "test",
+        )
+    )
+    assert resp.status_code == 200
+    payload = json.loads(bytes(resp.body).decode())
+    assert payload["data"][0]["embedding"] == [0.1, 0.2]
+    assert rt.embed_calls == before + 1
+    row = get_table("gateway_usage_ledger").find_one(owner_id="ledger_probe")
+    assert row is not None
+    assert row["backend"] == "embed"
+    assert row["endpoint"] == "embeddings"
+    assert row["requested_model"] == "molodetz~embed"
+    assert row["model"] == "qwen/qwen3-embedding-8b"
+    assert row["success"] == 1
+
+
 def test_embeddings_disabled_returns_503(local_db, monkeypatch):
     monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeEmbedClient_openai_gateway)
     svc = GatewayService()