Selfishly opinionated GitHub Copilot prompts, agents, instructions, and skills.
This repository also configures its own agents. See what my-awesome-copilot tells them →
JakeTRogers/my-awesome-copilot
Skill Claude CodeCodex
Generate Conventional Commit messages from provided staged Git context. Use when asked to write a commit message, summarize staged changes, classify a change as feat/fix/refactor/docs/test/build/ci/style/perf.
JakeTRogers/my-awesome-copilot
Skill Claude CodeCodex
Map uncommitted working-tree changes to the existing branch commits they belong to, then produce git commit --fixup and autosquash rebase commands for the user to run. Use when asked which commit a change belongs to, to plan fixup commits, to organize changes into fixups, to prepare an autosquash rebase, or after…
JakeTRogers/my-awesome-copilot
Skill Claude CodeCodex
Discover a repository and generate or refresh AGENTS.md as the canonical AI context file shared by GitHub Copilot and Claude Code, with CLAUDE.md as a pointer. Use when asked to document a repo for AI agents, build repo context, create or update AGENTS.md, onboard an agent to a codebase, or before starting a new…
JakeTRogers/my-awesome-copilot
Skill Claude CodeCodex
Inspect and evaluate GitHub PR review comments, respond with a clear technical position, and implement only valid feedback. Use when asked to review PR comments, respond to reviewers, address reviewer feedback, or make changes based on PR feedback.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: