The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses git commands locally to extract commit history and file differences. These operations are essential for the skill's functionality and are performed within the local repository environment.
[PROMPT_INJECTION]: The skill ingests commit messages which are external data. This risk of indirect prompt injection is addressed by a mandatory review step (Phase 7), where the user confirms the generated documentation before it is applied to the project.
[SAFE]: No indicators of malicious behavior, such as data exfiltration, credential theft, or remote code execution, were identified. The use of quoted heredocs and human verification gates demonstrates a security-conscious design.

docs-updater