interview-to-specs
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the repository's codebase and user-provided intent to generate documentation.
- Ingestion points: Reads existing specification files, ADRs, glossary files, and source code (SKILL.md Step 1).
- Boundary markers: Lacks explicit boundary markers or sanitization for repository content interpolated into generated files.
- Capability inventory: Performs local repository file read/write operations and executes shell commands for path validation.
- Sanitization: No specific sanitization or escaping for external content is specified.
- Note: The inclusion of a manual confirmation step ('Draft the spec, then confirm') provides a human-in-the-loop review mechanism.
- [COMMAND_EXECUTION]: The skill utilizes shell commands for repository management tasks.
- Evidence: Executes 'git check-ignore -q ' to verify that target file paths are not excluded by git configuration (SKILL.md Step 3).
Audit Metadata