utility-pm-critic
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's primary function is to provide a cross-client wrapper for a sub-agent (
pm-critic). It implements a "read and execute inline" pattern where it reads local documentation and agent files to guide its behavior. - [SAFE]: All references and dependencies trace back to the vendor's own infrastructure and local file paths (e.g.,
github.com/product-on-purpose/pm-skills). - [SAFE]: The skill uses conditional logic to determine the runtime environment (Claude Code vs. other clients), which is a legitimate architectural pattern for cross-platform compatibility.
- [SAFE]: No sensitive file access, network exfiltration, or obfuscation was found in the provided files. The instructions focus entirely on reviewing PM artifacts (PRDs, OKRs) against established standards.
Audit Metadata