feat(api-server): support X-Hermes-Session-Id header for session continuity

Allow callers to pass X-Hermes-Session-Id in request headers to continue
an existing conversation. When provided, history is loaded from SessionDB
instead of the request body, and the session_id is echoed in the response
header. Without the header, existing behavior is preserved (new uuid per
request).

This enables web UI clients to maintain thread continuity without modifying
any session state themselves — the same mechanism the gateway uses for IM
platforms (Telegram, Discord, etc.).

tests/gateway/test_api_server.py

@@ -1576,3 +1576,110 @@ class TestConversationParameter:
                 assert resp.status == 200
                 # Conversation mapping should NOT be set since store=false
                 assert adapter._response_store.get_conversation("ephemeral-chat") is None
+
+
+# ---------------------------------------------------------------------------
+# X-Hermes-Session-Id header (session continuity)
+# ---------------------------------------------------------------------------
+
+
+class TestSessionIdHeader:
+    @pytest.mark.asyncio
+    async def test_new_session_response_includes_session_id_header(self, adapter):
+        """Without X-Hermes-Session-Id, a new session is created and returned in the header."""
+        mock_result = {"final_response": "Hello!", "messages": [], "api_calls": 1}
+        app = _create_app(adapter)
+        async with TestClient(TestServer(app)) as cli:
+            with patch.object(adapter, "_run_agent", new_callable=AsyncMock) as mock_run:
+                mock_run.return_value = (mock_result, {"input_tokens": 0, "output_tokens": 0, "total_tokens": 0})
+                resp = await cli.post(
+                    "/v1/chat/completions",
+                    json={"model": "hermes-agent", "messages": [{"role": "user", "content": "Hi"}]},
+                )
+            assert resp.status == 200
+            assert resp.headers.get("X-Hermes-Session-Id") is not None
+
+    @pytest.mark.asyncio
+    async def test_provided_session_id_is_used_and_echoed(self, adapter):
+        """When X-Hermes-Session-Id is provided, it's passed to the agent and echoed in the response."""
+        mock_result = {"final_response": "Continuing!", "messages": [], "api_calls": 1}
+        app = _create_app(adapter)
+        async with TestClient(TestServer(app)) as cli:
+            with patch.object(adapter, "_run_agent", new_callable=AsyncMock) as mock_run, \
+                 patch("hermes_state.SessionDB") as mock_db_cls:
+                mock_db = MagicMock()
+                mock_db.get_messages_as_conversation.return_value = [
+                    {"role": "user", "content": "previous message"},
+                    {"role": "assistant", "content": "previous reply"},
+                ]
+                mock_db_cls.return_value = mock_db
+                mock_run.return_value = (mock_result, {"input_tokens": 0, "output_tokens": 0, "total_tokens": 0})
+
+                resp = await cli.post(
+                    "/v1/chat/completions",
+                    headers={"X-Hermes-Session-Id": "my-session-123"},
+                    json={"model": "hermes-agent", "messages": [{"role": "user", "content": "Continue"}]},
+                )
+
+            assert resp.status == 200
+            assert resp.headers.get("X-Hermes-Session-Id") == "my-session-123"
+            call_kwargs = mock_run.call_args.kwargs
+            assert call_kwargs["session_id"] == "my-session-123"
+
+    @pytest.mark.asyncio
+    async def test_provided_session_id_loads_history_from_db(self, adapter):
+        """When X-Hermes-Session-Id is provided, history comes from SessionDB not request body."""
+        mock_result = {"final_response": "OK", "messages": [], "api_calls": 1}
+        db_history = [
+            {"role": "user", "content": "stored message 1"},
+            {"role": "assistant", "content": "stored reply 1"},
+        ]
+        app = _create_app(adapter)
+        async with TestClient(TestServer(app)) as cli:
+            with patch.object(adapter, "_run_agent", new_callable=AsyncMock) as mock_run, \
+                 patch("hermes_state.SessionDB") as mock_db_cls:
+                mock_db = MagicMock()
+                mock_db.get_messages_as_conversation.return_value = db_history
+                mock_db_cls.return_value = mock_db
+                mock_run.return_value = (mock_result, {"input_tokens": 0, "output_tokens": 0, "total_tokens": 0})
+
+                resp = await cli.post(
+                    "/v1/chat/completions",
+                    headers={"X-Hermes-Session-Id": "existing-session"},
+                    # Request body has different history — should be ignored
+                    json={
+                        "model": "hermes-agent",
+                        "messages": [
+                            {"role": "user", "content": "old msg from client"},
+                            {"role": "assistant", "content": "old reply from client"},
+                            {"role": "user", "content": "new question"},
+                        ],
+                    },
+                )
+
+            assert resp.status == 200
+            call_kwargs = mock_run.call_args.kwargs
+            # History must come from DB, not from the request body
+            assert call_kwargs["conversation_history"] == db_history
+            assert call_kwargs["user_message"] == "new question"
+
+    @pytest.mark.asyncio
+    async def test_db_failure_falls_back_to_empty_history(self, adapter):
+        """If SessionDB raises, history falls back to empty and request still succeeds."""
+        mock_result = {"final_response": "OK", "messages": [], "api_calls": 1}
+        app = _create_app(adapter)
+        async with TestClient(TestServer(app)) as cli:
+            with patch.object(adapter, "_run_agent", new_callable=AsyncMock) as mock_run, \
+                 patch("hermes_state.SessionDB", side_effect=Exception("DB unavailable")):
+                mock_run.return_value = (mock_result, {"input_tokens": 0, "output_tokens": 0, "total_tokens": 0})
+
+                resp = await cli.post(
+                    "/v1/chat/completions",
+                    headers={"X-Hermes-Session-Id": "some-session"},
+                    json={"model": "hermes-agent", "messages": [{"role": "user", "content": "Hi"}]},
+                )
+
+            assert resp.status == 200
+            call_kwargs = mock_run.call_args.kwargs
+            assert call_kwargs["conversation_history"] == []
+            assert call_kwargs["session_id"] == "some-session"