cogworks-learn
Pass
Audited by Gen Agent Trust Hub on Mar 30, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill includes shell scripts
validate-skill.shandinstall-to-agents.shdesigned to automate skill verification and deployment within the local environment. - [EXTERNAL_DOWNLOADS]: The installation script references the
npx skillscommand, which fetches theskillsutility from the official npm registry at runtime. - [SAFE]: The skill incorporates robust quality gates and explicitly instructs the agent to treat user-provided design inputs as untrusted, reducing the surface for indirect prompt injection attacks.
- [SAFE]: Implements
disable-model-invocation: truein the frontmatter, a security control that prevents the agent from automatically triggering the skill without manual user intervention.
Audit Metadata