The Agent Skills Directory

[COMMAND_EXECUTION]: The workflow requires executing shell commands for Git operations such as git status, git add, git diff, and git commit to manage the working tree.
[COMMAND_EXECUTION]: The skill instructs the agent to identify and execute repository-specific "verification commands" (검증 명령) to validate changes before they are committed.
[PROMPT_INJECTION]: The skill processes untrusted data from the local repository, including file contents and history. This creates an Indirect Prompt Injection surface where malicious content embedded in the codebase could attempt to manipulate the agent's logic during the workflow.
Ingestion points: Reads data from git status --short, repository diffs, and existing commit history as specified in the input section.
Boundary markers: Absent; the skill does not define clear delimiters to separate untrusted repository data from its own core instructions.
Capability inventory: Includes file system modification (staging) and execution of arbitrary validation scripts found within the repository.
Sanitization: No explicit sanitization or filtering of the repository content is performed before processing.

to-commit