@companion-ai/feynman@0.5.40
The open source AI research agent.
AI Security Review
scanned 23d ago · by lpm-firewall-aiThe preinstall hook only checks the Node version, and project configuration writes occur through an explicit command that preserves existing files. The runtime’s disclosed, opt-out telemetry and signed-in research tools do not establish a concrete attack.
Decision evidence
public snapshot- package.json runs a Node version check at preinstall; the script exits only when the installed Node version is unsupported.
- The /init command writes AGENTS.md only when the user invokes it and the file does not already exist.
- bin/feynman.js checks for Pi dependencies and loads the package CLI; no credential transfer or hidden install action is shown there.
- dist/cli.js initializes telemetry and displays its notice; the notice describes the collected data and an opt-out setting.
- alpha.ts registers account-dependent research tools only when signed in; annotations are local.
- The telemetry notice states that prompts, model output, paper content, and tool arguments are not sent, and provides an opt-out.
Source & flagged code
6 flaggedPackage defines install-time lifecycle scripts.
package.jsonView on unpkgSource gates dangerous network, credential, or execution behavior behind CI, host, platform, time, or geo fingerprint checks.
bin/feynman.jsView on unpkg · L23Findings
Affected versions and remediation
This report applies to @companion-ai/feynman@0.5.40.
See all versions of @companion-ai/feynman for other recorded verdicts.
- Review the evidence and your use of @companion-ai/feynman@0.5.40 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.