10-andruia-skill-smith
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-defined goals and inputs to generate new instruction sets (SKILL.md files). This creates an ingestion surface where malicious user prompts could influence the behavior of the generated skill or the current agent during the validation phase.
- Ingestion points: User requests for skill creation or improvement (SKILL.md).
- Boundary markers: None explicitly specified for separating user-provided descriptions from the generated instructions.
- Capability inventory: File system write access (skills/ directory), execution of local npm scripts.
- Sanitization: No explicit sanitization or filtering of user-provided content is defined.
- [COMMAND_EXECUTION]: The workflow involves executing local repository commands to validate the generated skill.
- Evidence: Instructs the agent to run 'npm run chain', 'npm run validate:references', 'npm run security:docs', and 'npm run test'. These operations are scoped to the local environment and intended for testing integrity.
Audit Metadata