The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and execute instructions from external plan files, which constitutes an attack surface for indirect prompt injection.
Ingestion points: The agent is instructed to "Read plan file" and "Execution Readiness View" in the initial loading step (SKILL.md).
Boundary markers: The skill explicitly includes a manual verification step: "Review critically... Raise them [concerns] with your human partner before starting."
Capability inventory: The skill possesses the capability to perform Git operations (managing branches, worktrees, and commits) and read local files.
Sanitization: Security is mitigated through structured logic checks such as the TDD Route Guard, Change Necessity audit, and Pre-Edit Complexity Check, which force the agent to justify and validate steps before execution.

executing-plans