The Agent Skills Directory

[COMMAND_EXECUTION]: The skill automates several shell-based operations using git, gh (GitHub CLI), and a local Python script. Specifically, it is configured to perform a git push automatically after every commit without requesting user confirmation, which could lead to unintended code being published if the agent is misled by previous tasks.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it processes untrusted data from the local repository environment.
Ingestion points: The skill reads data from git remote get-url origin and commit message content via the .git/COMMIT_EDITMSG file.
Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the data being processed.
Capability inventory: The skill has the capability to perform write operations to the repository, including git commit, git push, gh pr create, and gh release create.
Sanitization: While scripts/validate_commit.py performs regex-based format validation, it does not sanitize the input for malicious instructions or shell-injection characters that could influence the agent's next steps.

git-workflow