@compr/opscontext-mcp@2.19.2
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 6d ago · by lpm-firewall-aiThe generated agent instruction can encourage broad commits and pushes, creating a concrete but user-triggered agent capability risk. The inspected manifest shows no automatic install hook, so this does not support a publish block.
Decision evidence
public snapshot- The CLI template tells an AI agent to commit and push all changed repositories; that instruction can encourage external changes when a user scaffolds a project.
- The manifest exposes CLI commands and defines a prepublish-only build script, with no automatic install hook shown.
- The inspected agent-integration locations concern project-file discovery, skill installation, hook settings, settings reads, and shared-index operation.
- The commit-and-push instruction appears in a CLI-generated template; the inspected package manifest does not configure it as an npm install-time action.
- The installer hooks are exposed through CLI entrypoints, and the inspected hook code adds OpsContext commands while preserving unrelated entries.
Source & flagged code
7 flagged · loading sourceThis package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/install-autostart.jsView on unpkgSource writes installer persistence such as shell profile or service configuration.
dist/install-autostart.jsView on unpkg · L13Findings
Affected versions and remediation
This report applies to @compr/opscontext-mcp@2.19.2.
See all versions of @compr/opscontext-mcp for other recorded verdicts.
- Review the evidence and your use of @compr/opscontext-mcp@2.19.2 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.