spec
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill implements a structured design process using interactive prompts and codebase exploration. The instructions focus on documentation and planning without suspicious behaviors.- [INDIRECT_PROMPT_INJECTION]: The skill processes external codebase content to map data models and services. This involves reading untrusted data (code files) which could potentially contain malicious instructions designed to influence agent behavior.
- Ingestion points: Codebase files and directory structure explored in Step 2.
- Boundary markers: The skill does not define explicit delimiters for codebase content.
- Capability inventory: File reading (codebase) and writing to the
.specs/directory. - Sanitization: The skill implements a specific slug-generation routine (lowercasing, removing filler words, hyphen-joining) for user-provided arguments used in filenames, which effectively prevents path traversal or injection into the file system path.
Audit Metadata