@metaphi-ai/hum@0.1.74
Hum: a self-improving coding agent for your terminal.
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThe automatic remote bootstrap and secondary package install are risky and insufficiently contained for a clean verdict. The source does not meet the stated threshold for malicious classification because no concrete theft, broad AI-agent control-surface mutation, or lifecycle self-dependency chain was found.
Decision evidence
public snapshot- The package runs a postinstall hook automatically during npm installation.
- That hook invokes an installer that downloads and executes uv's shell or PowerShell installer.
- The installer creates a home-directory environment and installs the external hum-cli Python package into it.
- It accepts an environment override for the Python package specification, widening the install-time payload source.
- No package source reads browser credentials, SSH keys, npm credentials, or environment secrets for transmission.
- The inspected JavaScript network connection is limited to localhost engine communication; no credential-exfiltration endpoint was found.
- No self-dependency on @metaphi-ai/hum or AI-agent configuration mutation was found in package.json or the installer.
Source & flagged code
12 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgFindings
Affected versions and remediation
This report applies to @metaphi-ai/hum@0.1.74.
See all versions of @metaphi-ai/hum for other recorded verdicts.
- Review the evidence and your use of @metaphi-ai/hum@0.1.74 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.