cheat-sheet-guide
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions are purely functional and oriented toward documentation generation. The skill does not attempt to execute code, exfiltrate data, or bypass safety controls. The use of frontmatter to disable model invocation is a positive security practice that limits the skill's reach.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from codebase files to identify project patterns, creating a minor injection surface. \n
- Ingestion points: Analyzes project-specific paths, configuration files, and code examples found within the local repository (SKILL.md, Step 2).\n
- Boundary markers: The instructions enforce a rigid markdown structure for the output, though they lack explicit delimiters to ignore embedded instructions within the source files.\n
- Capability inventory: The skill is limited to markdown generation and text output. It is explicitly restricted from making additional model calls via frontmatter configuration.\n
- Sanitization: No sanitization of the scanned codebase content is specified. Given the documentation use case and lack of dangerous capabilities, the risk is minimal.
Audit Metadata