pennyrouter@0.3.5
Install and manage PennyRouter local coding-agent integrations.
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThis is an explicit-user-command AI-agent integration with a real future-code execution risk from unpinned npx registration. It is not a concrete malicious npm install-hook chain because the manifest has no lifecycle hook.
Decision evidence
public snapshot- The explicit install command registers an MCP server in existing Claude Code and Codex configuration by default.
- The registered MCP command uses unpinned `npx -y pennyrouter mcp`, allowing a later agent session to fetch and run a registry release.
- The installer replaces Claude Code's status-line command with a package-written Node script.
- The installer can copy a runtime and append a managed PATH block to shell startup files.
- package.json has no preinstall, install, or postinstall hook.
- The configuration changes are reached through an explicit CLI install command, not package installation.
- CodeGPT setup only copies the supplied key to the clipboard and prints manual instructions; it does not alter IDE settings.
- Network calls are directed to the configured PennyRouter gateway for setup and credential storage; no unrelated exfiltration endpoint was found.
Source & flagged code
6 flagged · loading sourceSource writes installer persistence such as shell profile or service configuration.
src/launch.test.jsView on unpkg · L1Findings
Affected versions and remediation
This report applies to pennyrouter@0.3.5.
See all versions of pennyrouter for other recorded verdicts.
- Review the evidence and your use of pennyrouter@0.3.5 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.