irac-prompt-stephane-boghossian
Pass
Audited by Gen Agent Trust Hub on Jul 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security risks were detected in the skill instructions or metadata. The skill's functionality is limited to text restructuring and does not involve any privileged operations.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data (the 'raw ask') to generate structured output. While this represents a potential surface for indirect prompt injection, the risk is minimal because the skill lacks the capabilities to perform dangerous actions.
- Ingestion points: User-provided 'raw ask' processed via instructions in
SKILL.md. - Boundary markers: The skill is instructed to output the result inside a fenced code block to prevent the agent from accidentally executing the instructions.
- Capability inventory: No subprocess calls, file operations, or network requests are present in the skill.
- Sanitization: No explicit sanitization of user input is performed, but the output is intended for manual user copy-pasting or review.
Audit Metadata