The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill is susceptible to indirect prompt injection as it requires the agent to analyze the contents of git diffs to generate commit messages.
Ingestion points: untrusted data enters the context via git diff, git diff --staged, and git status --porcelain in SKILL.md.
Boundary markers: Absent. The skill does not instruct the agent to disregard instructions found within the code changes being analyzed.
Capability inventory: The skill has Bash tool access and executes git add and git commit commands.
Sanitization: Absent. There is no evidence of filtering or escaping the content of the diff before it is used to generate the commit message or CLI arguments.
[Command Execution] (SAFE): The skill uses Bash to execute standard git operations. It includes a 'Git Safety Protocol' that explicitly forbids destructive operations like force pushing or hard resets without user consent.

git-commit