cs-feat-design
Pass
Audited by Gen Agent Trust Hub on May 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute repository-specific Python tools (
.codestable/tools/validate-yaml.py,.codestable/tools/search-yaml.py) and the systemgreputility. These tools are used for document validation and information retrieval within the local codebase context. - [PROMPT_INJECTION]: The skill processes untrusted data from user inputs (slugs, summaries) and repository files (roadmap items, existing design documents). This constitutes an indirect prompt injection surface.
- Ingestion points: User input via chat (feature slugs, requirement summaries), and local repository files including
.codestable/features/, roadmap files, and source code. - Boundary markers: The skill provides strict templates in
reference.mdand instructs the agent to use specific Markdown headers and YAML structures, reducing the risk of accidental instruction following. - Capability inventory: The skill possesses the ability to read and write files within the project directory and execute local scripts and CLI tools.
- Sanitization: The workflow incorporates a mandatory validation step using
validate-yaml.pyto ensure the integrity of generated configuration files.
Audit Metadata