The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a Python script (scripts/inspect_repository.py) to interface with the Git CLI. The script employs subprocess.Popen with argument lists (avoiding shell=True), which is a secure practice to prevent shell injection vulnerabilities when handling user-provided paths.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze data from Git repositories, such as branch names, remotes, and status information. To mitigate potential risks from malicious repository metadata, the instructions emphasize read-only observation and require explicit user authorization for any proposed modifications. It also provides clear guidelines on distinguishing between repository facts and AI suggestions.
[CREDENTIALS_UNSAFE]: The execution environment is hardened in scripts/inspect_repository.py by stripping sensitive Git environment variables (e.g., GIT_CONFIG_PARAMETERS, GIT_DIR) and disabling interactive prompts (GIT_TERMINAL_PROMPT=0). This prevents the accidental capture or leakage of user credentials during repository inspection.
[EXTERNAL_DOWNLOADS]: The skill references authoritative documentation from well-known and trusted providers including GitHub, GitLab, Microsoft, and the Git Project. These references are used for informational purposes and do not involve the execution of remote code.

git-awesome