From git-intent-interface
Process the last N commits as instructions — execute !! markers and commit message directives in any file type. This skill should be used when the user says "apply intent", "process last N commits", "execute markers", "run git intent", "apply !! instructions", or "g-intent-apply".
How this skill is triggered — by the user, by Claude, or both
Slash command
/git-intent-interface:g-intent-applyThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
If `$ARGUMENTS` is empty or not a positive integer, ask the user for the number of commits.
If $ARGUMENTS is empty or not a positive integer, ask the user for the number of commits.
Verify the working tree is clean (git status --porcelain returns empty). If not, stop and ask the user to commit or stash first.
Process the last $ARGUMENTS commits as instructions:
Run git log -p -$ARGUMENTS --reverse and display the output
Read all changed files in full (skip binary files)
Interpret the instructions:
!! — the text after !! is a global instruction for the committed files!! alone — no global instruction, just process the markers in files!!) — the message is context/explanation, not an instruction; still process any !! markers in the files!! markers in files — specific instructions tied to their locationExecute all instructions. For each !! marker:
!!!!)[keep], preserve it as-isCommit each logical change separately (one commit per marker processed or per coherent edit from the commit message) with a descriptive subject and mandatory body explaining the reasoning
Markers use the native comment syntax of the file's language, prefixed with !!.
When removing a marker, remove the entire comment (including closing delimiters like */ or -->), not just the !! token.
!! markers removed (except [keep]), code/text is valid and clean!! are preserved — they are persistent context, not instructionsnpx claudepluginhub maxim-uvarov/my-claude-skills --plugin git-intent-interfaceAnalyzes uncommitted changes, groups by cohesion, and generates commit messages matching project style. Outputs git commands or executes with --execute.
Manages Git commit workflow using Conventional Commits format with safety protocols. Creates, validates, executes commits; handles hooks, PRs, and safety checks before operations.
Guides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.