optimize-prompt
Pass
Audited by Gen Agent Trust Hub on Mar 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No evidence of prompt injection or instructions to bypass safety guidelines was found. The skill maintains a helpful and instructional tone consistent with its stated purpose.\n- [SAFE]: The skill uses the 'Read' tool exclusively for accessing internal reference files (references/system-prompts.md and references/agentic-prompts.md) within the skill's own directory. There is no access to sensitive system files or external network resources.\n- [SAFE]: No obfuscation, remote code execution, or persistence mechanisms are present in the skill's instructions or metadata.
Audit Metadata