effective-agent-skills
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The static analysis identified phrases such as "ignore previous instructions" within the authoring guide. Evaluation confirms these are part of a security checklist designed to teach users how to audit third-party skills for risks. The text is educational and does not attempt to override the agent's system instructions.
- [SAFE]: The skill is purely instructional and does not request access to sensitive files, network resources, or external tools. It functions as a cognitive discipline guide for structured workflow development.
- [EXTERNAL_DOWNLOADS]: The skill references external repositories (David Ondrej and Cursor) and documentation sites (agentskills.io) for attribution and further reading. No automated downloads or execution of remote content occur.
- [METADATA_POISONING]: The name, description, and workflow instructions are consistent with the stated purpose of helping users author and debug agent skills.
Audit Metadata