Skip to content

fix: bump mcp/fastmcp minimums for CVEs#144

Closed
zerone0x wants to merge 1 commit intotirth8205:mainfrom
zerone0x:fix/fastmcp-cves-139
Closed

fix: bump mcp/fastmcp minimums for CVEs#144
zerone0x wants to merge 1 commit intotirth8205:mainfrom
zerone0x:fix/fastmcp-cves-139

Conversation

@zerone0x
Copy link
Copy Markdown
Contributor

@zerone0x zerone0x commented Apr 8, 2026

Summary\n- bump mcp to >=2.0.0,<3\n- bump fastmcp to >=2.14.0,<3 to avoid critical CVEs\n\nFixes #139\n\n## Test plan\n- not run (dependency version change only)\n\n🤖 Generated with Claude Code

…Authored-By: Claude <noreply@anthropic.com>
@tirth8205
Copy link
Copy Markdown
Owner

Closing as duplicate of #143 (same change from the same author). We'll merge #143 instead. Thanks!

@tirth8205
Copy link
Copy Markdown
Owner

This PR addresses the fastmcp CVE dependency issue (#139) which is still not resolved in main. The current pyproject.toml still has mcp>=1.0.0,<2 which may resolve to vulnerable fastmcp versions. Worth reviving — it's a 2-line change to update dependency constraints.

@tirth8205
Copy link
Copy Markdown
Owner

Thank you for this @zerone0x — closing as superseded by PR #222, which bumps `fastmcp` to `>=2.14.0,<3`. That PR covers the same CVE fix plus the transitive `fakeredis` breakage (#195), the Windows event-loop hang (#46 / #136), and several unrelated fixes. Verified locally: all 24 tools register on fastmcp 2.14.6, all 695 tests pass, coverage 73.6%.

Really appreciate you flagging this — you were first on the CVE and your PR pushed the conversation forward. #222 will ship in the next release (v2.2.4) once Windows testing lands.

@tirth8205 tirth8205 closed this Apr 11, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants