agent-md-refactor
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is a purely instructional guide for documentation management. It defines a workflow for analyzing, categorizing, and restructuring text-based configuration files without the use of scripts or external tools.
- [INDIRECT_PROMPT_INJECTION]: The skill processes existing agent instruction files which constitute an external data ingestion surface.
- Ingestion points: Processes content from files such as
AGENTS.md,CLAUDE.md, andCOPILOT.md(SKILL.md). - Boundary markers: No explicit delimiters or boundary markers are instructed for use during the ingestion or output phases.
- Capability inventory: The skill contains no code or tool definitions; it relies on the agent's native capabilities for file reading and creation.
- Sanitization: No specific sanitization or escaping of the ingested instruction content is defined.
Audit Metadata