The Agent Skills Directory

[SAFE]: The skill's functionality is entirely consistent with its described purpose of automating complex Git workflows. It adheres to security best practices by utilizing backup branches and requesting user consent before history-rewriting operations.
[COMMAND_EXECUTION]: The skill utilizes legitimate shell commands via git and gh to inspect repository state, manage branches, and sync changes. Operations like git push --force-with-lease are used correctly to prevent accidental overwriting of remote work, and all execution steps are transparently presented to the user.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted commit messages and pull request metadata.
Ingestion points: Commit messages are read via git log and PR details are fetched through gh pr view.
Boundary markers: The skill does not use specific delimiters to wrap the content of ingested commit messages.
Capability inventory: The skill can manipulate local and remote repository states through branch resets and pushes.
Sanitization: The logic incorporates --fixed-strings when searching commit messages to prevent accidental character interpretation.

vp-stacked-pr-rebase