The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill establishes an ingestion surface for untrusted data by instructing the agent to analyze git diff and git diff --staged output as described in SKILL.md. While no explicit boundary markers or sanitization methods are defined for this data, the agent uses this context solely to generate commit messages and branch names. This represents a potential surface for codebase content to influence agent behavior, but the risk is low and inherent to the intended developer-assistance use case.
[COMMAND_EXECUTION]: The agent workflow involves executing standard, non-privileged Git commands (git status, git diff) to gather context from the repository. These operations are performed locally and are strictly necessary for the skill's purpose of automating structured commit message creation.

commits