trace-mcp@3.22.0
Framework-aware code intelligence MCP server — 87 framework integrations, 81 languages, 90.6% fewer input tokens to review a pull request
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThis is a real, automatic lifecycle attack surface rather than a scanner-only signal. The behavior is package-aligned but still dangerous because it persists a daemon and replaces executable application code from a mutable remote release channel.
Decision evidence
public snapshot- The automatic postinstall chain runs two scripts that alter user-level state.
- On macOS, postinstall writes and bootstraps a persistent LaunchAgent with RunAtLoad and KeepAlive.
- The installer fetches a latest GitHub release and replaces installed application bundles with its downloaded archive.
- The checksum is downloaded from the same release source as the archive, so it does not establish an independent trust boundary.
- The remote update path is limited to macOS and targets trace-mcp application bundles.
- The code has opt-out environment variables and checks a daemon-disabled marker.
- No credential harvesting or secret exfiltration was found in the inspected lifecycle scripts.
Source & flagged code
16 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgFindings
Affected versions and remediation
This report applies to trace-mcp@3.22.0.
See all versions of trace-mcp for other recorded verdicts.
- Review the evidence and your use of trace-mcp@3.22.0 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.