[BUG] 2.1.9 regression: Uncontrolled request spam to /mcp and OAuth endpoints by Claude Code

Status Fixed / completed
Reported on v2.1.9
Maintainer reply ✓ Yes — ashwin-ant
Activity 5 comments · opened Jan 16, 2026 · closed Jan 17, 2026
💡 Likely answer: A maintainer (ashwin-ant, collaborator) responded on this thread — see the highlighted reply below.

Preflight Checklist

  • [x] I have searched existing issues and this hasn't been reported yet
  • [x] This is a single bug report (please file separate reports for different bugs)
  • [x] I am using the latest version of Claude Code

What's Wrong?

We operate Context7, Claude Code 2.1.9 is sending an extremely excessive number of requests to our /mcp and /oauth endpoints, effectively DDoS-ing our infrastructure. This is not a minor performance issue — our DDoS protection has been triggered multiple times, causing service degradation for our legitimate users.
The timestamp of the Claude Code 2.1.9 release correlates exactly with a sudden, massive spike in traffic to our servers. This is causing real operational harm to our platform.

We've checked and the timestamp of the release of Claude Code 2.1.9 matches exactly with the increase in the load of our servers.

What Should Happen?

Claude Code should make requests to MCP and OAuth endpoints only when necessary — not thousands of requests in a matter of minutes. Whatever polling, retry logic, or connection mechanism was changed in 2.1.9 needs immediate attention.

Error Messages/Logs

<img width="1453" height="626" alt="Image" src="https://github.com/user-attachments/assets/d417086e-c4f7-47cf-b082-57b7d4d53dee" />

You can see the request counts in the last 30 minutes.
<img width="1479" height="672" alt="Image" src="https://github.com/user-attachments/assets/601a8f84-4ec6-4314-affe-35b68acfd176" />

Affected endpoints:

  • /mcp
  • /.well-known/oauth-authorization-server/mcp
  • /.well-known/openid-configuration/mcp
  • /mcp/.well-known/openid-configuration

Steps to Reproduce

Not directly reproducable, but we can see from our firewall requests that some clients sends thousands of requests in few minutes.

Claude Model

Not sure / Multiple models

Is this a regression?

Yes, this worked in a previous version

Last Working Version

2.1.8

Claude Code Version

2.1.9

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

From our firewall and server logs:

Multiple Claude Code 2.1.9 user agents are sending thousands of requests within minutes
This traffic pattern is indistinguishable from a DDoS attack
Our rate limiting and DDoS protection have been triggered, impacting legitimate users
This began exactly when 2.1.9 was released and did not occur with 2.1.8

We urge you to treat this as a high-priority issue. MCP server operators should not have their infrastructure overwhelmed by the client implementation. Please consider a hotfix or rollback of whatever change caused this regression.

View original on GitHub ↗

5 Comments

nicku33 · 7 months ago

This is also happening to our orgs MCP server. I had 4 Claude tabs alive and was spamming our servers with 0-50 OAuth requests per second. Effectively a DDoS

ashwin-ant collaborator · 7 months ago

2.1.11 is out with a fix for this issue!

nicku33 · 7 months ago
2.1.11 is out with a fix for this issue!

Thanks for the patch @ashwin-ant . Just to confirm do we need to only block 2.1.9 or 2.1.10 as well ?

ashwin-ant collaborator · 7 months ago

2.1.10 also contains the bug, but it's unlikely many users updated to it. Blocking both would be safest though, yeah.

github-actions[bot] · 7 months ago

This issue has been automatically locked since it was closed and has not had any activity for 7 days. If you're experiencing a similar issue, please file a new issue and reference this one if it's relevant.