craft
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input (skill ideas or conversation history) to generate new instructions. It lacks explicit boundary markers or sanitization steps to prevent the user from injecting malicious instructions into the generated skill's body.
- Ingestion points: User-provided skill ideas or conversation history processed in SKILL.md.
- Boundary markers: Not present.
- Capability inventory: The skill can potentially use file-writing capabilities if explicitly requested by the user during the packaging step.
- Sanitization: Not present.
- [NO_CODE]: The skill consists entirely of markdown instructions and YAML frontmatter. It does not include any executable scripts, binaries, or shell commands.
Audit Metadata