gentle-ai-skill-improver
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill audits and refactors instructions in other
SKILL.mdfiles, creating a data ingestion surface where content from those files could influence the agent's execution. - Ingestion points: Reads workspace
SKILL.mdfiles,docs/skill-style-guide.md, and.atl/skill-registry.md. - Boundary markers: The instructions do not define delimiters or specific markers to isolate the audited content from the active skill's logic.
- Capability inventory: The skill performs file system read and write operations to audit and modify local files.
- Sanitization: No explicit logic is defined to sanitize or escape instructions found within the audited files.
Audit Metadata