oauth for tools

docs/configuration.mdx

@@ -123,6 +123,49 @@ mcp-agent uses two configuration files:
   </Tab>
 </Tabs>
 
+## OAuth Configuration
+
+MCP Agent exposes two complementary OAuth configuration blocks:
+
+- `authorization` describes how the MCP Agent server validates inbound bearer tokens and publishes protected resource metadata.
+- `oauth` configures delegated authorization when the agent connects to downstream MCP servers.
+
+```yaml
+authorization:
+  enabled: true
+  issuer_url: https://auth.example.com
+  resource_server_url: https://agent.example.com/mcp
+  required_scopes: ["mcp.read", "mcp.write"]
+  introspection_endpoint: https://auth.example.com/oauth/introspect
+  introspection_client_id: ${INTROSPECTION_CLIENT_ID}
+  introspection_client_secret: ${INTROSPECTION_CLIENT_SECRET}
+
+oauth:
+  callback_base_url: https://agent.example.com
+  flow_timeout_seconds: 180
+  token_store:
+    backend: memory   # use "redis" for multi-instance deployments
+    refresh_leeway_seconds: 90
+
+mcp:
+  servers:
+    github:
+      transport: streamable_http
+      url: https://github.mcp.example.com/mcp
+      auth:
+        oauth:
+          enabled: true
+          scopes: ["repo", "user:email"]
+          client_id: ${GITHUB_MCP_CLIENT_ID}
+          client_secret: ${GITHUB_MCP_CLIENT_SECRET}
+          redirect_uri_options:
+            - https://agent.example.com/internal/oauth/callback
+```
+
+- When `authorization.enabled` is true the MCP server advertises `/.well-known/oauth-protected-resource` and enforces bearer tokens using the provided introspection or JWKS configuration.
+- `oauth` enables delegated authorization flows; the default in-memory token store is ideal for local development while Redis is recommended for production clusters.
+- Downstream servers opt into OAuth via `mcp.servers.<name>.auth.oauth`. Supplying a `client_id`/`client_secret` allows immediate usage; support for dynamic client registration is planned as a follow-up.
+
 ## Configuration Reference
 
 ### Execution Engine

docs/oauth_support_design.md

@@ -0,0 +1,108 @@
+# MCP Agent OAuth Support
+
+## Goals
+- Protect MCP Agent Cloud servers using OAuth 2.1 so MCP clients obtain tokens via standard flows.
+- Enable MCP Agent runtimes to authenticate to downstream MCP servers that require OAuth access tokens.
+- Provide pluggable token storage for both local development (in-memory) and multi-instance deployments (Redis planned).
+- Maintain compatibility with MCP Authorization spec (RFC 8414, RFC 9728, OAuth 2.1 + PKCE, Resource Indicators) and the proposed delegated authorization SEP.
+
+## Architecture Overview
+
+### Components
+1. **Auth Server Integration** – Configure the FastMCP instance with `AuthSettings` and a custom `TokenVerifier` that calls MCP Agent Cloud auth services.
+2. **Protected Resource Metadata** – Serve `/.well-known/oauth-protected-resource` using FastMCP hooks so clients can discover the auth server.
+3. **Access Token Validation** – Enforce bearer tokens on every inbound MCP request via `RequireAuthMiddleware`, populating the request context with the authenticated user.
+4. **OAuth Token Service** – New `mcp_agent.oauth` package with:
+   - `TokenStore`/`TokenRecord` abstractions
+   - `InMemoryTokenStore` and Redis-backed implementation (second pass)
+   - `TokenManager` orchestration (acquire, refresh, revoke)
+   - `OAuthHttpxAuth` for attaching tokens to downstream HTTP transports
+   - `AuthorizationFlowCoordinator` that interacts with the user via MCP `auth/request`
+5. **Delegated Authorization UI Flow** – Extend the gateway/session relay so servers can send `auth/request` messages to MCP clients, capturing authorization codes via either:
+   - Client-returned callback URL (preferred, works with SEP-capable clients)
+   - MCP Agent hosted callback endpoint (`/internal/oauth/callback/{flow_id}`) as a fallback / native-app style loopback.
+6. **Configuration Surface** – Extend `Settings` and per-server `MCPServerAuthSettings` to describe OAuth behaviour (scopes, preferred auth server, redirect URIs, etc.) and global token-store configuration.
+
+### Key Data Flow
+1. **Inbound Requests**
+   - Client presents bearer token ⇒ `BearerAuthBackend` + `MCPAgentTokenVerifier` introspect token.
+   - Verified token populates context with `OAuthUserIdentity` (provider + subject + email).
+   - Context is propagated into workflows/sessions so downstream OAuth flows know the acting user.
+
+2. **Outbound HTTP (downstream MCP server)**
+   - `ServerRegistry` detects `auth.oauth` configuration.
+   - Wraps HTTP transport with `OAuthHttpxAuth` which requests an access token from `TokenManager`.
+   - `TokenManager` checks store; if missing/expired ⇒ `AuthorizationFlowCoordinator` performs RFC 9728 discovery, PKCE, delegated browser flow through MCP client, exchanges code for tokens, caches result.
+   - Requests automatically retry after token refresh when a response returns 401/invalid token.
+
+3. **Token Storage**
+   - Tokens stored per `(user_identity, resource, authorization_server)` tuple with metadata (scopes, expiry, refresh token, provider claims).
+   - Store implements optimistic locking to avoid concurrent refresh storms.
+   - Pluggable backend (`InMemoryTokenStore` initial, Redis follow-up).
+
+## Module Plan
+
+```
+src/mcp_agent/oauth/
+  __init__.py
+  identity.py           # OAuthUserIdentity, helpers to extract from auth context
+  records.py            # TokenRecord dataclass/pydantic model
+  store/base.py         # TokenStore protocol
+  store/in_memory.py    # Default store
+  manager.py            # TokenManager (get/refresh/invalidate)
+  flow.py               # AuthorizationFlowCoordinator
+  http/auth.py          # OAuthHttpxAuth (httpx.Auth implementation)
+  metadata.py           # RFC 8414 + RFC 9728 discovery helpers
+  pkce.py               # PKCE + state utilities
+  errors.py             # Custom exception hierarchy
+```
+
+Integration touchpoints:
+- `mcp_agent/config.py` – add OAuth settings models.
+- `mcp_agent/core/context.py` – add `current_user`, `token_manager`, `token_store`, `oauth_config` fields.
+- `mcp_agent/app.py` – initialize token store/manager based on settings.
+- `mcp_agent/server/app_server.py` – configure FastMCP auth settings, register callback route, surface user identity, extend relay to handle `auth/request`.
+- `mcp_agent/mcp/mcp_server_registry.py` & `mcp_agent/mcp/mcp_connection_manager.py` – wire `OAuthHttpxAuth` into HTTP transports and expose helper for manual token teardown.
+- `mcp_agent/mcp/client_proxy.py` – add proxy helpers for `auth/request`.
+- `SessionProxy` – add direct request helper for `auth/request` and ensure Temporal flow support.
+- `examples/mcp_agent_server/*` – demonstrate configuration changes.
+- Tests – new suite exercising token store, metadata discovery, flow orchestration (with mocked HTTP + client responses).
+
+## OAuth Flow Details
+1. **Discovery**
+   - If downstream server responds 401 with `WWW-Authenticate`, parse for `resource_metadata` ⇒ GET metadata ⇒ determine auth server URL(s).
+   - Fetch authorization server metadata (RFC 8414).
+   - Perform optional dynamic client registration when configured and supported.
+
+2. **Authorization Request**
+   - Generate PKCE challenge/verifier, secure `state`, choose `redirect_uri`.
+   - Build authorization URL including `resource` parameter (RFC 8707) + requested scopes.
+   - Invoke `auth/request` via SessionProxy → MCP client opens browser.
+
+3. **Callback Handling**
+   - Preferred: MCP client returns callback URL payload via request result.
+   - Fallback: Authorization server redirects to `/internal/oauth/callback/{flow_id}`.
+   - Coordinator validates `state`, extracts `code` (and errors).
+
+4. **Token Exchange / Storage**
+   - POST token endpoint with code + PKCE verifier + resource.
+   - Store access token, refresh token, expiry, scope, provider metadata.
+   - Associate tokens with user identity for reuse.
+
+5. **Refresh / Revocation**
+   - Manager refreshes when expiry within configurable grace window.
+   - Invalidate token on refresh failure or when server responses indicate revocation.
+   - Provide method to revoke tokens via authorization server when supported.
+
+## Open Questions / Follow-ups
+- Redis-backed `TokenStore` (requires deployment secrets) – planned second phase.
+- How LastMile auth server exposes token introspection + JWKS; need concrete endpoint specs to finalize `MCPAgentTokenVerifier`.
+- MCP client adoption of `auth/request` SEP – need capability detection; until widely supported we rely on hosted callback fallback & manual instructions.
+- Access control DSL (include/exclude by email/domain) – to be evaluated once token identity payload finalized.
+
+## Testing Strategy
+- Unit tests for token store concurrency + expiry handling.
+- Metadata discovery + PKCE generation (pure python tests).
+- Integration-style test for delegated flow using mocked HTTP server + fake MCP client (ensures `auth/request` plumbing works end-to-end).
+- Tests around server 401 enforcement + WWW-Authenticate header.
+

examples/oauth/README.md

@@ -0,0 +1,19 @@
+# OAUTH scenarios
+
+## preconfigured
+
+In this case, a token is hard-coded into the configuration. 
+This is useful for testing or when the token is static.
+
+## workflow_pre_auth
+
+In this case, the client can call a `workflows_pre_auth` tool before calling a workflow to seed the tokens.
+This is useful when the client can do the auth step, but the workflow cannot (e.g. because it runs async).
+There is a slight hack employed here: since we don't have oauth for the mcp app, we do not have a user.
+Since we need a user to store the token against, we create a synthetic user and use that.
+
+## dynamic_auth
+
+In this case, no tokens are provided, and the calls comes back to the client to do the auth step.
+Currently implemented as an elicitation request (to align with the future elicit URL scheme).
+I have not achieved full end-to-end flow here.

examples/oauth/dynamic_auth/client.py

@@ -0,0 +1,163 @@
+import asyncio
+import time
+
+from datetime import timedelta
+from anyio.streams.memory import MemoryObjectReceiveStream, MemoryObjectSendStream
+from mcp import ClientSession
+from mcp.types import CallToolResult, LoggingMessageNotificationParams
+from mcp_agent.app import MCPApp
+from mcp_agent.config import MCPServerSettings
+from mcp_agent.core.context import Context
+from mcp_agent.mcp.gen_client import gen_client
+from mcp_agent.mcp.mcp_agent_client_session import MCPAgentClientSession
+from mcp_agent.human_input.console_handler import console_input_callback
+from mcp_agent.elicitation.handler import console_elicitation_callback
+
+from rich import print
+
+try:
+    from exceptiongroup import ExceptionGroup as _ExceptionGroup  # Python 3.10 backport
+except Exception:  # pragma: no cover
+    _ExceptionGroup = None  # type: ignore
+try:
+    from anyio import BrokenResourceError as _BrokenResourceError
+except Exception:  # pragma: no cover
+    _BrokenResourceError = None  # type: ignore
+
+async def main():
+    # Create MCPApp to get the server registry
+    app = MCPApp(
+        name="workflow_mcp_client",
+        human_input_callback=console_input_callback,
+        elicitation_callback=console_elicitation_callback,
+    )
+    async with app.run() as client_app:
+        logger = client_app.logger
+        context = client_app.context
+
+        # Connect to the workflow server
+        logger.info("Connecting to workflow server...")
+
+        # Override the server configuration to point to our local script
+        context.server_registry.registry["basic_agent_server"] = MCPServerSettings(
+            name="basic_agent_server",
+            description="Local workflow server running the basic agent example",
+            transport="sse",
+            url="http://127.0.0.1:8000/sse",
+        )
+
+        # Define a logging callback to receive server-side log notifications
+        async def on_server_log(params: LoggingMessageNotificationParams) -> None:
+            level = params.level.upper()
+            name = params.logger or "server"
+            print(f"[SERVER LOG] [{level}] [{name}] {params.data}")
+
+        # Provide a client session factory that installs our logging callback
+        # and prints non-logging notifications to the console
+        class ConsolePrintingClientSession(MCPAgentClientSession):
+            async def _received_notification(self, notification):  # type: ignore[override]
+                try:
+                    method = getattr(notification.root, "method", None)
+                except Exception:
+                    method = None
+
+                # Avoid duplicating server log prints (handled by logging_callback)
+                if method and method != "notifications/message":
+                    try:
+                        data = notification.model_dump()
+                    except Exception:
+                        data = str(notification)
+                    print(f"[SERVER NOTIFY] {method}: {data}")
+
+                return await super()._received_notification(notification)
+
+        def make_session(
+            read_stream: MemoryObjectReceiveStream,
+            write_stream: MemoryObjectSendStream,
+            read_timeout_seconds: timedelta | None,
+            context: Context | None = None,
+        ) -> ClientSession:
+            return ConsolePrintingClientSession(
+                read_stream=read_stream,
+                write_stream=write_stream,
+                read_timeout_seconds=read_timeout_seconds,
+                logging_callback=on_server_log,
+                context=context,
+            )
+
+        try:
+            async with gen_client(
+                "basic_agent_server",
+                context.server_registry,
+                client_session_factory=make_session,
+            ) as server:
+                try:
+                    await server.set_logging_level("info")
+                except Exception:
+                    # Older servers may not support logging capability
+                    print("[client] Server does not support logging/setLevel")
+
+                # List available tools
+                tools_result = await server.list_tools()
+                logger.info(
+                    "Available tools:",
+                    data={"tools": [tool.name for tool in tools_result.tools]},
+                )
+
+                print(await server.call_tool("github_org_search", {"query": "lastmileai"}))
+        except Exception as e:
+            # Tolerate benign shutdown races from stdio client (BrokenResourceError within ExceptionGroup)
+            if _ExceptionGroup is not None and isinstance(e, _ExceptionGroup):
+                subs = getattr(e, "exceptions", []) or []
+                if (
+                    _BrokenResourceError is not None
+                    and subs
+                    and all(isinstance(se, _BrokenResourceError) for se in subs)
+                ):
+                    logger.debug("Ignored BrokenResourceError from stdio shutdown")
+                else:
+                    raise
+            elif _BrokenResourceError is not None and isinstance(
+                e, _BrokenResourceError
+            ):
+                logger.debug("Ignored BrokenResourceError from stdio shutdown")
+            elif "BrokenResourceError" in str(e):
+                logger.debug(
+                    "Ignored BrokenResourceError from stdio shutdown (string match)"
+                )
+            else:
+                raise
+        # Nudge cleanup of subprocess transports before the loop closes to avoid
+        # 'Event loop is closed' from BaseSubprocessTransport.__del__ on GC.
+        try:
+            await asyncio.sleep(0)
+        except Exception:
+            pass
+        try:
+            import gc
+
+            gc.collect()
+        except Exception:
+            pass
+
+
+def _tool_result_to_json(tool_result: CallToolResult):
+    if tool_result.content and len(tool_result.content) > 0:
+        text = tool_result.content[0].text
+        try:
+            # Try to parse the response as JSON if it's a string
+            import json
+
+            return json.loads(text)
+        except (json.JSONDecodeError, TypeError):
+            # If it's not valid JSON, just use the text
+            return None
+
+
+if __name__ == "__main__":
+    start = time.time()
+    asyncio.run(main())
+    end = time.time()
+    t = end - start
+
+    print(f"Total run time: {t:.2f}s")