@constellationdev/mcp@1.3.0
Codebase Understanding for AI Coding Agents
AI Security Review
scanned 3d ago · by lpm-firewall-aiThe package sends default-on runtime telemetry containing project and branch identifiers and tool usage details, without disclosing that reporting in its README. The opt-out makes this a suspicious telemetry finding rather than a malicious exfiltration finding.
Decision evidence
public snapshot- Usage tracking defaults to enabled, with environment-variable opt-outs.
- After a successful code_intel execution, the package builds a usage event from project, branch, invocation, and timing data.
- The event includes the project ID, branch name, tool invocation names, token counts, and timing metrics.
- The package POSTs the event with the configured access key to its usage endpoint.
- The README describes the package as providing code intelligence without transmitting source code.
- The only lifecycle script is prepare: husky; no install or postinstall telemetry hook is declared.
- The runtime telemetry can be disabled with CONSTELLATION_USAGE_METRICS=false or USAGE_TRACKING_ENABLED=false.
Source & flagged code
5 flaggedThis package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/index.jsView on unpkgFindings
Affected versions and remediation
This report applies to @constellationdev/mcp@1.3.0.
See all versions of @constellationdev/mcp for other recorded verdicts.
- Review the evidence and your use of @constellationdev/mcp@1.3.0 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.