gitrama_skill

Overview

This skill brings AI-powered semantic search and intelligence to Git history. It embeds an AskGIT assistant that inspects commits, diffs, branches, and contributors to answer natural-language questions and generate contextual outputs like commit messages and branch names. All AI analysis is done server-side via api.gitrama.ai with no API keys required.

How this skill works

The CLI gathers repository context using a set of git subprocess calls (commits, diffs, branch info, file tree and contributors) and posts that context to the gitrama API for analysis. AskGIT supports multi-turn conversations by maintaining a short conversation history and returning concise, structured answers. Separate commands generate conventional commit messages from staged changes and produce descriptive branch names from task descriptions.

When to use it

• You need a quick, human-readable summary of recent commits or what changed in a module.
• You want AI-generated conventional commit messages for staged diffs.
• You need concise, standardized branch names from a task or ticket description.
• You want contributor insights, change-frequency analysis, or release summaries.
• You are doing code archaeology and need to know who changed what and why.

Best practices

• Run the tool inside a git repository so the CLI can gather full context.
• Stage only the changes you want summarized before generating commit messages.
• Provide clear task descriptions when requesting branch names for better suggestions.
• Review generated messages/names before committing to ensure project-specific conventions are respected.
• Use gtr chat iteratively to refine answers; AskGIT keeps recent conversation context for follow-ups.

Example use cases

• Ask: 'Summarize the last 20 commits' to get a concise change log for a release note.
• Run gtr commit to produce a conventional commit message for your staged changes.
• Run gtr branch 'add user authentication with OAuth' to get a clean branch name like feature/auth-oauth.
• Ask: 'Who has been the main contributor to the API layer?' to get contributor activity and attribution.
• Investigate breaking changes between two tags by asking AskGIT to analyze diffs and flagged risky edits.