@compr/opscontext-mcp@2.8.1
OpsContext for AI Agents — read-only fleet visibility (PM2/nginx/Docker/git/cron) + tamper-evident audit log + policy-as-code hooks. The ops + compliance layer Claude Code can't grow natively.
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThis is not a malicious install hook because no install lifecycle trigger exists, but explicit commands modify an AI-agent control surface and create persistence. The audit-discouraging source comment adds a concrete social-engineering signal.
Decision evidence
public snapshot- Activation source contains a comment telling auditors not to review it, which is reviewer manipulation.
- An explicit CLI command installs Claude Code hooks and rewrites its settings file.
- The installed hook reads Claude hook input, including prompts and tool-call previews, then sends it to a local collector.
- An explicit CLI command creates and bootstraps a persistent macOS LaunchAgent.
- package.json has only a prepublishOnly script, not an install-time lifecycle hook.
- The Claude-hook and autostart behaviors are dispatched only by named CLI commands.
- The hook posts to 127.0.0.1 rather than an external endpoint.
- No runtime dependency on this package itself is declared.
Source & flagged code
7 flagged · loading sourcePackage source references dynamic require/import behavior.
dist/adapters.jsView on unpkg · L46A manifest entrypoint or package-local install chain reaches persistence behavior.
dist/install-autostart.jsView on unpkg · L13Findings
Affected versions and remediation
This report applies to @compr/opscontext-mcp@2.8.1.
See all versions of @compr/opscontext-mcp for other recorded verdicts.
- Review the evidence and your use of @compr/opscontext-mcp@2.8.1 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.