The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes bash commands including awk, git, grep, and test to perform deterministic validation of local files such as plan.md. These commands are used to verify section presence, requirement tagging, and decision rationale completeness.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted local data to drive autonomous file modifications in 'Fix Mode'. Ingestion points: Untrusted content from plan.md, checks.yaml, and codebase artifacts is read during the assessment process. Boundary markers: There are no explicit markers or instructions defining the boundary between the skill's logic and the data being processed. Capability inventory: The skill can execute shell commands and perform autonomous file system writes when the 'fix' argument is provided. Sanitization: The skill does not implement sanitization or validation of the input data's intent before applying autonomous fixes.

confidence