agent-md-refactor
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from user-provided agent instruction files which may contain instructions that could influence the agent during the refactoring process. However, since the skill lacks dangerous capabilities like network access or arbitrary command execution, the risk is minimal.
- Ingestion points: The agent reads the contents of
CLAUDE.md,AGENTS.md, andCOPILOT.mdfiles throughout the 5-phase refactoring process. - Boundary markers: The skill does not define specific boundaries or "ignore instruction" markers for the content being refactored.
- Capability inventory: The skill involves file system operations (read and write) to create the
.claude/directory and reorganized markdown files. - Sanitization: The skill does not sanitize the contents of the instruction files before processing them.
Audit Metadata