fix: use smart defaults for Gateway Mode client_id

CHANGELOG.md

@@ -7,6 +7,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [1.7.1] - 2026-01-25
 
+### Changed
+
+- **Gateway Mode smart defaults**: `get_policy_approved_context()` and `audit_llm_call()` now use `"community"` as default client_id when not configured, enabling zero-config usage for community/self-hosted deployments
+
 ### Fixed
 
 - **PolicyCategory**: Added `PII_SINGAPORE = "pii-singapore"` enum value for Singapore PII detection policies (NRIC, FIN, UEN patterns)

axonflow/client.py

@@ -394,25 +394,17 @@ def _has_credentials(self) -> bool:
         """
         return bool(self._config.client_id)
 
-    def _require_credentials(self, feature: str) -> None:
-        """Require credentials for enterprise features.
+    def _get_effective_client_id(self) -> str:
+        """Get the effective client_id, using smart default for community mode.
 
-        Raises AuthenticationError if client_id is not configured.
-        Note: client_secret is optional for community mode.
+        Returns the configured client_id if set, otherwise returns "community"
+        as a smart default. This enables zero-config usage for community/self-hosted
+        deployments while still supporting enterprise deployments with explicit credentials.
 
-        Args:
-            feature: Name of the feature requiring credentials (for error message)
-
-        Raises:
-            AuthenticationError: If client_id is not configured
+        Returns:
+            The client_id to use in requests
         """
-        if not self._has_credentials():
-            msg = (
-                f"{feature} requires client_id. "
-                "Set client_id when creating the client "
-                "(client_secret is optional for community mode)."
-            )
-            raise AuthenticationError(msg)
+        return self._config.client_id if self._config.client_id else "community"
 
     async def __aenter__(self) -> AxonFlow:
         """Async context manager entry."""
@@ -1160,8 +1152,8 @@ async def get_policy_approved_context(
         LLM call to ensure policy compliance.
 
         Note:
-            This is an enterprise feature that requires credentials.
-            Set client_id and client_secret when creating the client.
+            Uses smart default "community" for client_id if not configured,
+            enabling zero-config usage for community/self-hosted deployments.
 
         Args:
             user_token: JWT token for the user making the request
@@ -1173,7 +1165,7 @@ async def get_policy_approved_context(
             PolicyApprovalResult with context ID and approved data
 
         Raises:
-            AuthenticationError: If credentials are not configured or user token is invalid
+            AuthenticationError: If user token is invalid
             ConnectionError: If unable to reach AxonFlow Agent
             TimeoutError: If request times out
 
@@ -1186,12 +1178,12 @@ async def get_policy_approved_context(
             >>> if not result.approved:
             ...     raise PolicyViolationError(result.block_reason)
         """
-        # Gateway Mode is an enterprise feature that requires credentials
-        self._require_credentials("Gateway Mode (get_policy_approved_context)")
+        # Use smart default for client_id - enables zero-config community mode
+        client_id = self._get_effective_client_id()
 
         request_body = {
             "user_token": user_token,
-            "client_id": self._config.client_id,
+            "client_id": client_id,
             "query": query,
             "data_sources": data_sources or [],
             "context": context or {},
@@ -1322,12 +1314,12 @@ async def audit_llm_call(
             ...     latency_ms=250
             ... )
         """
-        # Gateway Mode is an enterprise feature that requires credentials
-        self._require_credentials("Gateway Mode (audit_llm_call)")
+        # Use smart default for client_id - enables zero-config community mode
+        client_id = self._get_effective_client_id()
 
         request_body = {
             "context_id": context_id,
-            "client_id": self._config.client_id,
+            "client_id": client_id,
             "response_summary": response_summary,
             "provider": provider,
             "model": model,

tests/test_auth_headers.py

@@ -133,67 +133,89 @@ async def test_no_auth_headers_for_health_check(self, httpx_mock):
 
 
 class TestEnterpriseFeatureValidation:
-    """Test that enterprise features require client_id before making requests."""
+    """Test that Gateway Mode uses smart defaults for client_id."""
 
     @pytest.mark.asyncio
-    async def test_pre_check_fails_without_client_id(self, httpx_mock):
-        """get_policy_approved_context should fail before making request when no client_id."""
-        # Don't mock the endpoint - we should fail before making the request
+    async def test_pre_check_uses_smart_default_without_client_id(self, httpx_mock):
+        """get_policy_approved_context should use 'community' as default client_id."""
+        httpx_mock.add_response(
+            url="http://localhost:8080/api/policy/pre-check",
+            json={
+                "context_id": "ctx_smart_default",
+                "approved": True,
+                "policies": [],
+                "expires_at": "2025-12-20T12:00:00Z",
+            },
+        )
+
         client = AxonFlow(
             endpoint="http://localhost:8080",
-            # No client_id - truly no credentials
+            # No client_id - should use "community" smart default
             debug=True,
         )
 
         async with client:
-            with pytest.raises(AuthenticationError) as exc_info:
-                await client.get_policy_approved_context(
-                    user_token="",
-                    query="Test query",
-                )
+            result = await client.get_policy_approved_context(
+                user_token="test-user",
+                query="Test query",
+            )
 
-            assert "requires client_id" in str(exc_info.value)
-            assert "Gateway Mode" in str(exc_info.value)
+            assert result.approved is True
+            assert result.context_id == "ctx_smart_default"
 
-        # No request should have been made
+        # Verify request was made with smart default client_id
         requests = httpx_mock.get_requests()
-        assert len(requests) == 0
+        assert len(requests) == 1
+        import json
+
+        body = json.loads(requests[0].content)
+        assert body["client_id"] == "community"
 
-        print("✅ get_policy_approved_context fails without client_id (no request made)")
+        print("✅ get_policy_approved_context uses 'community' smart default")
 
     @pytest.mark.asyncio
-    async def test_audit_fails_without_client_id(self, httpx_mock):
-        """audit_llm_call should fail before making request when no client_id."""
-        # Don't mock the endpoint - we should fail before making the request
+    async def test_audit_uses_smart_default_without_client_id(self, httpx_mock):
+        """audit_llm_call should use 'community' as default client_id."""
+        httpx_mock.add_response(
+            url="http://localhost:8080/api/audit/llm-call",
+            json={
+                "success": True,
+                "audit_id": "audit_smart_default",
+            },
+        )
+
         client = AxonFlow(
             endpoint="http://localhost:8080",
-            # No client_id - truly no credentials
+            # No client_id - should use "community" smart default
             debug=True,
         )
 
         async with client:
-            with pytest.raises(AuthenticationError) as exc_info:
-                await client.audit_llm_call(
-                    context_id="ctx_123",
-                    response_summary="Test response",
-                    provider="openai",
-                    model="gpt-4",
-                    token_usage=TokenUsage(
-                        prompt_tokens=100,
-                        completion_tokens=50,
-                        total_tokens=150,
-                    ),
-                    latency_ms=250,
-                )
-
-            assert "requires client_id" in str(exc_info.value)
-            assert "Gateway Mode" in str(exc_info.value)
-
-        # No request should have been made
+            result = await client.audit_llm_call(
+                context_id="ctx_123",
+                response_summary="Test response",
+                provider="openai",
+                model="gpt-4",
+                token_usage=TokenUsage(
+                    prompt_tokens=100,
+                    completion_tokens=50,
+                    total_tokens=150,
+                ),
+                latency_ms=250,
+            )
+
+            assert result.success is True
+            assert result.audit_id == "audit_smart_default"
+
+        # Verify request was made with smart default client_id
         requests = httpx_mock.get_requests()
-        assert len(requests) == 0
+        assert len(requests) == 1
+        import json
+
+        body = json.loads(requests[0].content)
+        assert body["client_id"] == "community"
 
-        print("✅ audit_llm_call fails without client_id (no request made)")
+        print("✅ audit_llm_call uses 'community' smart default")
 
     @pytest.mark.asyncio
     async def test_pre_check_works_with_credentials(self, httpx_mock):