prompt-library-curator
Pass
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes untrusted user data (pasted prompts or external files). However, it implements security best practices by wrapping all user content within markdown code blocks in the output, which serves as a boundary marker to prevent embedded instructions from affecting downstream agent actions.\n- [COMMAND_EXECUTION]: The skill performs standard file system operations by reading from user-specified text or markdown files and writing the final catalog to 'prompt-library.md'. These operations are restricted to the local workspace and are essential to the skill's primary function.\n- [DATA_EXFILTRATION]: No network operations or external data transfer mechanisms were detected. All processing and output occur within the agent's local environment.
Audit Metadata