The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a local Python script (scripts/collect_release_context.py) to gather repository information. This script executes standard Git commands such as git log, git diff, and git branch. Analysis of the script shows it uses subprocess.run() with a list of arguments and shell=False (default), which is the recommended secure practice to prevent shell command injection.
[DATA_EXPOSURE]: The skill accesses Git commit history and file diffs. While these are processed by the agent, the data remains local to the repository environment, and no network exfiltration patterns were detected. This access is necessary for the skill's primary function of summarizing changes.
[PROMPT_INJECTION]: As the skill processes untrusted input in the form of Git commit messages (Indirect Prompt Injection surface), there is a theoretical risk that malicious instructions in a commit message could influence the agent's output. However, the instructions provide clear formatting rules and a specific workflow that limits the agent's behavior to writing release metadata files, making this a low-risk scenario.

helmor-release