The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the local repository.
Ingestion points: The output of git diff is injected directly into the prompt in SKILL.md via dynamic context injection.
Boundary markers: No delimiters or isolation markers are used to separate the diff content from the agent's instructions.
Capability inventory: The skill explicitly instructs the agent to "proceed with creating the commits," giving it the authority to perform file system writes (git commits) based on the processed content.
Sanitization: No sanitization or filtering is applied to the diff output before it is processed by the agent.
[DYNAMIC_CONTEXT_INJECTION]: The skill uses the !command`` syntax in SKILL.md to execute date, git status, and git diff. These are standard developer tools used here for information gathering and are consistent with the skill's primary purpose.

commit