trace-mcp@3.36.1
Framework-aware code intelligence MCP server — 88 framework integrations, 81 languages, 72.7% fewer input tokens to review a pull request, comprehension at parity
AI Security Review
scanned 12d ago · by lpm-firewall-aiSource inspection found package-owned setup and ordinary development scripts, with no concrete attack behavior. The static similarity and agent-location hints did not establish malicious behavior.
Decision evidence
public snapshot- package.json defines a postinstall chain for native preflight and package-owned app/control-plane setup.
- The agent-location hits in dist/extract-worker.js and dist/proxy.js are only paths to Claude user data; no configuration write was found at those locations.
- The benchmark reads local Claude session files; it iterates session logs and records evaluation results.
- The flagged scripts describe screenshots, upgrade checks, fixture recording, throughput/performance benchmarks, documentation generation, and response-token measurement.
- No concrete malicious behavior was identified in the inspected source.
- The package runs a postinstall hook that can perform local setup and updater work.
Source & flagged code
26 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@3.36.1.
See all versions of trace-mcp for other recorded verdicts.
- Review the evidence and your use of trace-mcp@3.36.1 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.