The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes JSON artifacts from 'memory/artifacts/prompt-enhancer/' that contain user-influenced requirements, representing a surface for indirect prompt injection. \n
Ingestion points: The 'Artifact Passing' section in SKILL.md describes reading requirements from external files. \n
Boundary markers: The skill uses JSON formatting but does not include explicit instructions for the agent to ignore instructions embedded within the requirement data. \n
Capability inventory: The skill is restricted to Read, Write, and AskUserQuestion tools. \n
Sanitization: Artifact content is used directly without mentioned validation or escaping. \n- [COMMAND_EXECUTION]: Documentation and examples in SKILL.md provide shell commands like ls, grep, and ln for the agent to manage and identify implementation artifacts. \n- [SAFE]: The skill includes a mandatory plan-review phase via AskUserQuestion, ensuring that all generated plans undergo human validation before the workflow proceeds.

planning