The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local Python scripts (scripts/write_prd.py and scripts/validate_prd.py) included in the repository to process requirements and validate the output. These scripts use standard library modules and perform intended file system operations to save the generated PRD artifacts.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8). It collects free-text responses from users and interpolates them directly into Markdown files that are intended to be consumed by other AI agents.
Ingestion points: User responses gathered via the AskUserQuestion tool.
Boundary markers: Content is organized using Markdown headings and bullet points, but there are no explicit instructions or delimiters to prevent a downstream agent from following malicious instructions embedded in the user's requirements.
Capability inventory: The skill performs file writes to the local disk and instructs the agent to perform Git operations for wiki publishing.
Sanitization: The skill does not perform any escaping or validation of user input before rendering it into the PRD templates.
[SAFE]: The skill implements a security best practice by automatically redacting sensitive evaluation criteria and scoring rubrics from the version of the PRD intended for the improvement agent (PRD.improver_only.md), preventing evaluation leakage and overfitting.

skillhone-prd