The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the subprocess module in scripts/selftest.py to run its internal validation scripts against test fixtures and in scripts/workflow_route.py to execute git rev-parse HEAD. These operations are limited to the skill's own code and local version control, providing no pathway for arbitrary command injection.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-generated markdown specifications, which creates an attack surface for indirect prompt injection.
Ingestion points: Data is read from plan, check, and verification markdown files in the .specs/ folder.
Boundary markers: The skill mandates strict templates, specific requirement IDs, and EARS (Easy Approach to Requirements Syntax).
Capability inventory: The agent can modify project files and execute local Python validators.
Sanitization: The skill employs deterministic Python validators and requires manual human review of plans and checks before proceeding, which mitigates the risk of instructions embedded in data.

wtk-lean