slap
Pass
Audited by Gen Agent Trust Hub on May 12, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill employs authoritative directives such as "You must refine your environment" and "You are not permitted to proceed until you prove to me you can operate this way," which are common patterns in behavioral override and jailbreak attempts.
- [PROMPT_INJECTION]: It mandates the agent to make "meta-level changes to the repo, docs, and your own behavior" based on user input (steering), which effectively allows a user to permanently alter the agent's operating constraints and repository content through natural language.
- [NO_CODE]: The skill contains no executable code or external dependencies, consisting purely of markdown instructions.
Audit Metadata