agent-instruction-files
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary purpose is to have the agent analyze, relocate, and prune instructions derived from various project files.
- Ingestion points: The instructions require the agent to read internal documentation files (e.g., AGENTS.md, CLAUDE.md), source code, and inline comments across the repository to identify guidance worth preserving or relocating.
- Boundary markers: The skill does not specify the use of delimiters or 'ignore' instructions when processing external content, which could lead the agent to follow instructions embedded in comments or documentation it is supposed to be auditing.
- Capability inventory: The agent is granted the capability to modify core instruction files (AGENTS.md, CLAUDE.md), which define the agent's behavior and constraints.
- Sanitization: There are no explicit instructions to sanitize, escape, or validate the content being moved into higher-privilege instruction files.
Audit Metadata