agent-config-tidy
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from user-provided directories, including agent documentation and scripts. This presents a surface for indirect prompt injection where instructions embedded in the target files could attempt to influence the agent's logic during the tidying process.
- Ingestion points: Recursively scans target directories for files like AGENTS.md, soul.md, SKILL.md, YAML configs, and various scripts (.py, .sh, .js, .ts, .ps1, etc.).
- Boundary markers: The skill explicitly mandates a 'Diff then Confirm' workflow, requiring user approval for each change, and creates .bak files for recovery.
- Capability inventory: Performs directory scanning, file reading, and file writing to rewrite configuration content and script comments.
- Sanitization: No automated sanitization is performed; the skill relies on the user's manual review of the generated difference report as the primary safety control.
Audit Metadata