The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes uncommitted changes via git diff, creating a potential surface for indirect prompt injection if file contents contain malicious instructions. \n
Ingestion points: The skill reads uncommitted file contents using git diff and git diff --cached (SKILL.md). \n
Boundary markers: The skill includes a mandatory confirmation step where the agent must present a staging plan to the user for approval before creating commits (SKILL.md, Step 4). \n
Capability inventory: The agent is authorized to use git add and git commit to modify the repository (SKILL.md, Step 6). \n
Sanitization: No specific sanitization or filtering of diff content is described in the instructions. \n- [COMMAND_EXECUTION]: The skill uses the Bash tool to execute git commands such as status, diff, merge-base, add, and commit. These commands are necessary for the skill's primary function and are explicitly restricted in the allowed-tools frontmatter field in SKILL.md to ensure they are only used for git-related operations.

commit