The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute ls commands to verify the existence of upstream artifacts (e.g., ls docs/01_BRD/ docs/02_PRD/). This is used for workflow validation and does not access sensitive system paths.
[COMMAND_EXECUTION]: The skill utilizes local bash and Python scripts (e.g., prd_core_wrapper_hook.sh, validate_prd.py, validate_cross_document.py) to perform document validation and quality checks. These scripts are referenced via relative paths within the project structure.
[PROMPT_INJECTION]: The skill contains an inherent attack surface by processing data from external files (Layer 1 BRD artifacts). It instructs the agent to read these files and extract information holistically. While there are no explicit boundary markers to prevent indirect prompt injection from those files, the capabilities of the skill are limited to file system read/write and validation within the project scope.
[EXTERNAL_DOWNLOADS]: No remote downloads or network operations (e.g., curl, wget) are present in the skill instructions. All referenced templates and validation tools are expected to be available locally within the ai_dev_ssd_flow or project directories.

doc-prd