psyclaw@0.30.6
A standalone, evidence-grounded social-science research agent with research contracts, evidence gates, and recoverable workflows
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThis is not malicious by the install-hook policy because it has no registry-install lifecycle trigger or demonstrated theft/destruction. Automatic mutation of an agent dependency at launch warrants a warning.
Decision evidence
public snapshot- The user-invoked executable automatically rewrites files in the installed Pi dependency, including its package metadata, interactive mode, settings, commands, tools, and footer.
- The manifest registers Pi extensions and bundled skills for an agent runtime.
- package.json has no preinstall, install, or postinstall hook; its only lifecycle hook is prepublishOnly.
- The bundled Python file flagged as a hidden payload is ordinary pytest fixture configuration.
- Panel telemetry imports are conditional on a supplied DSN or project key and otherwise return without network activity.
- No source evidence shows secret exfiltration, destructive behavior, or an automatic self-dependency chain.
Source & flagged code
5 flaggedPackage source references dynamic require/import behavior.
dist/apps/panel/observability.jsView on unpkg · L39Findings
Affected versions and remediation
This report applies to psyclaw@0.30.6.
See all versions of psyclaw for other recorded verdicts.
- Review the evidence and your use of psyclaw@0.30.6 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.