pennyrouter@0.2.29
Install and manage PennyRouter local coding-agent integrations.
AI Security Review
scanned 11d ago · by lpm-firewall-aiSource inspection found user-invoked Claude agent configuration and a status-line request authenticated with a PennyRouter key. This supports a warning for an agent capability, but the inspected source does not establish a malicious attack or automatic npm install behavior.
Decision evidence
public snapshot- The CLI install path configures Claude Code’s status line and writes a PennyRouter settings overlay.
- The status line reads a PennyRouter key and sends it to the configured base URL, so its behavior depends on the endpoint in the user’s settings.
- The executable is an explicit CLI command; the manifest shows no npm lifecycle hook.
- The inspected Codex, MCP, and thread-session target locations concern test fixtures or local session paths, not agent-control writes.
- CodeGPT setup prints manual instructions and does not write JetBrains settings.
- The package manifest exposes a CLI and defines only start and check scripts; no install-time hook is declared.
- The status line fetch is gated on a session ID, base URL, and PennyRouter key, and targets the configured gateway’s status endpoint.
- Claude installation preserves an existing status line by wrapping it, and the user invokes the install flow to configure the integration.
Source & flagged code
6 flagged · loading sourceSource writes installer persistence such as shell profile or service configuration.
src/launch.test.jsView on unpkg · L28Source gates dangerous network, credential, or execution behavior behind CI, host, platform, time, or geo fingerprint checks.
Findings
Affected versions and remediation
This report applies to pennyrouter@0.2.29.
See all versions of pennyrouter for other recorded verdicts.
- Review the evidence and your use of pennyrouter@0.2.29 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.