The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted data from Git history and remote branches which represents an indirect prompt injection surface. Ingestion points: Reads output from git log, git diff, and git pull as described in SKILL.md. Boundary markers: The instructions do not define delimiters to separate Git content from agent instructions. Capability inventory: Includes git command execution and file system write access. Sanitization: No filtering or validation of commit messages or pulled content is defined.
[SAFE]: The skill incorporates security best practices by explicitly instructing the agent to identify and exclude sensitive files such as credentials, .env files, and private keys from the staging process.
[COMMAND_EXECUTION]: The skill executes standard Git CLI commands to perform its core functions.
[SAFE]: The workflow requires explicit user confirmation before performing a git push, ensuring the user remains in control of remote changes and preventing automated exfiltration to remote repositories.
[SAFE]: The skill implements a 'Safe commit' protocol that uses temporary files to handle multi-line messages, avoiding shell-specific injection risks and ensuring cross-platform stability.

git-commit