-
Notifications
You must be signed in to change notification settings - Fork 2.6k
Implement RFC9728 - Support WWW-Authenticate header by MCP client #1071
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Implement RFC9728 - Support WWW-Authenticate header by MCP client #1071
Conversation
Hi, Any updates on it ? @ihrpr we rely on correct parsing of Auth header |
@ihrpr Can we raise attention to this bug? It blocks the MCP client from correctly finding the resource metadata, leading to incorrect AS. Thank you |
src/mcp/client/auth.py
Outdated
refresh_request = await self._refresh_token() | ||
refresh_response = yield refresh_request | ||
if response.status_code == 401: | ||
if self.context.can_refresh_token(): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm not sure I agree with this part... if we know our current tokens are invalid, and we know we can refresh them, shouldn't we do that before sending a request we know is going to 401?
I think that would mean moving this if/else
block ahead of line 526 where we yield request.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thank you for the suggestion @pcarleton
I was hesitant about it, so I implemented the "try first" approach, as it seemed more reliable.
As this article states, it's also perfectly fine to have that kind of optimization and refresh the token preemptively, so I updated the code as you proposed.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
Thanks for the thorough PR with tests, and being open to feedback!
Note to self: we'll still need to follow up to fix the path-based URL issue #1052
* Add regression test for stateless request memory cleanup (modelcontextprotocol#1140) * Implement RFC9728 - Support WWW-Authenticate header by MCP client (modelcontextprotocol#1071) * Add streamable HTTP starlette example to Python SDK docs (modelcontextprotocol#1111) * fix markdown error in README in main (modelcontextprotocol#1147) * README - replace code snippets with examples - add lowlevel to snippets (modelcontextprotocol#1150) * README - replace code snippets with examples - streamable http (modelcontextprotocol#1155) * chore: don't allow users to create issues outside the templates (modelcontextprotocol#1163) * Tests(cli): Add coverage for helper functions (modelcontextprotocol#635) * Docs: Update CallToolResult parsing in README (modelcontextprotocol#812) Co-authored-by: Felix Weinberger <[email protected]> * docs: add pre-commit install guide on CONTRIBUTING.md (modelcontextprotocol#995) Co-authored-by: Felix Weinberger <[email protected]> * fix flaky fix-test_streamablehttp_client_resumption test (modelcontextprotocol#1166) * README - replace code snippets with examples -- auth examples (modelcontextprotocol#1164) * Support falling back to OIDC metadata for auth (modelcontextprotocol#1061) * Add CODEOWNERS file for sdk (modelcontextprotocol#1169) * fix flaky test test_88_random_error (modelcontextprotocol#1171) * Make sure `RequestId` is not coerced as `int` (modelcontextprotocol#1178) * Fix: Replace threading.Lock with anyio.Lock for Ray deployment compatibility (modelcontextprotocol#1151) * fix: fix OAuth flow request object handling (modelcontextprotocol#1174) * update codeowners group (modelcontextprotocol#1191) * fix: perform auth server metadata discovery fallbacks on any 4xx (modelcontextprotocol#1193) * server: skip duplicate response on CancelledError (modelcontextprotocol#1153) Co-authored-by: ihrpr <[email protected]> * Unpack settings in FastMCP (modelcontextprotocol#1198) * chore: Remove unused prompt_manager.py file (modelcontextprotocol#1229) Co-authored-by: Tapan Chugh <[email protected]> * Improved supported for ProtectedResourceMetadata (modelcontextprotocol#1235) Co-authored-by: Paul Carleton <[email protected]> * chore: Remove unused variable notification_options (modelcontextprotocol#1238) * Improve README around the Context object (modelcontextprotocol#1203) * fix: allow to pass `list[str]` to `token_endpoint_auth_signing_alg_values_supported` (modelcontextprotocol#1226) * Remove strict validation on `response_modes_supported` member of `OAuthMetadata` (modelcontextprotocol#1243) * Add pyright strict mode on the whole project (modelcontextprotocol#1254) * Consistent casing for default headers Accept and Content-Type (modelcontextprotocol#1263) * Update dependencies and fix type issues (modelcontextprotocol#1268) Co-authored-by: Marcelo Trylesinski <[email protected]> * fix: prevent async generator cleanup errors in StreamableHTTP transport (modelcontextprotocol#1271) Co-authored-by: David Soria Parra <[email protected]> * chore: uncomment .idea/ in .gitignore (modelcontextprotocol#1287) Co-authored-by: Claude <[email protected]> * docs: clarify streamable_http_path configuration when mounting servers (modelcontextprotocol#1172) * feat: Add CORS configuration for browser-based MCP clients (modelcontextprotocol#1059) Co-authored-by: Marcelo Trylesinski <[email protected]> Co-authored-by: Felix Weinberger <[email protected]> * Added Audio to FastMCP (modelcontextprotocol#1130) * fix: avoid uncessary retries in OAuth authenticated requests (modelcontextprotocol#1206) Co-authored-by: Felix Weinberger <[email protected]> * Add PATHEXT to default STDIO env vars in windows (modelcontextprotocol#1256) * fix: error too many values to unpack (expected 2) (modelcontextprotocol#1279) Signed-off-by: San Nguyen <[email protected]> Co-authored-by: Felix Weinberger <[email protected]> Co-authored-by: Felix Weinberger <[email protected]> * SDK Parity: Avoid Parsing Server Response for non-JsonRPCMessage Requests (modelcontextprotocol#1290) * types: Setting default value for method: Literal (modelcontextprotocol#1292) * changes structured temperature to not deadly (modelcontextprotocol#1328) * Update simple-resource example to use non-deprecated read_resource return type (modelcontextprotocol#1331) Co-authored-by: Claude <[email protected]> * docs: Update README to include link to API docs for modelcontextprotocol#1329 (modelcontextprotocol#1330) * Allow ping requests before initialization (modelcontextprotocol#1312) * Python lint: Ruff rules for pylint and code complexity (modelcontextprotocol#525) * Fix context injection for resources and prompts (modelcontextprotocol#1336) * fix(fastmcp): propagate mimeType in resource template list (modelcontextprotocol#1186) Co-authored-by: Felix Weinberger <[email protected]> * fix: allow elicitations accepted without content (modelcontextprotocol#1285) Co-authored-by: Olivier Schiavo <[email protected]> * Use --frozen in pre-commit config (modelcontextprotocol#1375) * Return HTTP 403 for invalid Origin headers (modelcontextprotocol#1353) * Add test for ProtectedResourceMetadataParsing (modelcontextprotocol#1236) Co-authored-by: Paul Carleton <[email protected]> Co-authored-by: Marcelo Trylesinski <[email protected]> Co-authored-by: Felix Weinberger <[email protected]> * Fastmcp logging progress example (modelcontextprotocol#1270) Co-authored-by: Felix Weinberger <[email protected]> * feat: add paginated list decorators for prompts, resources, and tools (modelcontextprotocol#1286) Co-authored-by: Claude <[email protected]> * Remove "unconditionally" from conditional description (modelcontextprotocol#1289) * Use streamable-http consistently in examples (modelcontextprotocol#1389) * feat: Add SDK support for SEP-1034 default values in elicitation schemas (modelcontextprotocol#1337) Co-authored-by: Tapan Chugh <[email protected]> Co-authored-by: Felix Weinberger <[email protected]> * Implementation of SEP 973 - Additional metadata + icons support (modelcontextprotocol#1357) * Merge upstream/main with custom filtering --------- Signed-off-by: San Nguyen <[email protected]> Co-authored-by: Felix Weinberger <[email protected]> Co-authored-by: yurikunash <[email protected]> Co-authored-by: Pamela Fox <[email protected]> Co-authored-by: Inna Harper <[email protected]> Co-authored-by: Marcelo Trylesinski <[email protected]> Co-authored-by: Ian Davenport <[email protected]> Co-authored-by: Dagang Wei <[email protected]> Co-authored-by: Felix Weinberger <[email protected]> Co-authored-by: Stanley Law <[email protected]> Co-authored-by: Luca Chang <[email protected]> Co-authored-by: leweng <[email protected]> Co-authored-by: Clare Liguori <[email protected]> Co-authored-by: lukacf <[email protected]> Co-authored-by: ihrpr <[email protected]> Co-authored-by: Tapan Chugh <[email protected]> Co-authored-by: Tapan Chugh <[email protected]> Co-authored-by: Yann Jouanin <[email protected]> Co-authored-by: Paul Carleton <[email protected]> Co-authored-by: Sreenath Somarajapuram <[email protected]> Co-authored-by: Omer Korner <[email protected]> Co-authored-by: joesavage-silabs <[email protected]> Co-authored-by: Gregory L <[email protected]> Co-authored-by: David Soria Parra <[email protected]> Co-authored-by: Moustapha Ebnou <[email protected]> Co-authored-by: Max Isbey <[email protected]> Co-authored-by: Claude <[email protected]> Co-authored-by: Jerome <[email protected]> Co-authored-by: xavier <[email protected]> Co-authored-by: keurcien <[email protected]> Co-authored-by: Tim Esler <[email protected]> Co-authored-by: San Nguyen <[email protected]> Co-authored-by: Justin Wang <[email protected]> Co-authored-by: jess <[email protected]> Co-authored-by: Peter Alexander <[email protected]> Co-authored-by: Reid Geyer <[email protected]> Co-authored-by: Eleftheria Stein-Kousathana <[email protected]> Co-authored-by: Christian Clauss <[email protected]> Co-authored-by: pchoudhury22 <[email protected]> Co-authored-by: owengo <[email protected]> Co-authored-by: Olivier Schiavo <[email protected]> Co-authored-by: Steve Billings <[email protected]> Co-authored-by: Mike Salvatore <[email protected]>
Resolves #1054
This PR implements proper WWW-Authenticate header parsing for protected resource metadata URL discovery as required by RFC9728 and the MCP specification. This change ensures full compliance with MCP authentication requirements and includes several related improvements to the authentication flow.
Changes Made
Primary Changes
Indirect Improvements
Motivation and Context
As per MCP specification:
This implementation was missing from the current MCP client, creating a gap in specification compliance that has real-world implications. The absence of proper WWW-Authenticate header handling prevents hosting multiple MCP servers that share the same domain but function as separate OAuth resources.
How Has This Been Tested?
The changes were verified by the comprehensive unit-test coverage for different scenarios, as well as simulating the Authorization scenario by executing the client locally.
Breaking Changes
Authorization Flow Change: The client now follows proper OAuth flow by sending an initial request to the resource and only performing authorization upon receiving a 401 status code. This aligns with OAuth specifications and security best practices.
Impact: Clients that relied on the previous behavior (where authorization was attempted regardless of initial response) may need to be updated. However, this change improves security and compliance with OAuth standards.
Migration: Most properly configured MCP servers should work without changes. Servers that expect immediate authorization without sending a 401 response should be updated to follow OAuth specifications.
Types of changes
Checklist
Additional context