The Agent Skills Directory

[SAFE]: The skill demonstrates safe operational practices by requiring explicit human-in-the-loop confirmation before executing commands and instructing the agent to avoid broad staging commands like git add ..
[INDIRECT_PROMPT_INJECTION]: The skill identifies a potential injection surface associated with processing code diffs and session history, which is managed via human review: 1. Ingestion points: Reads git status, git diff, and conversation history (SKILL.md). 2. Boundary markers: Requires the agent to present a plan to the user for approval before execution (Step 3, SKILL.md). 3. Capability inventory: Utilizes git add, git commit, and pre-commit run (SKILL.md). 4. Sanitization: Execution is gated by human review of the drafted plan.

commit