Code review, evaluation, and fix utilities for quality assurance workflows
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Multi-agent collaboration plugin suite for AI coding tools — task-driven workflow via tmux, code review, and learning aids.
ai-kit provides a collection of plugins that coordinate multiple AI agents (Claude Code, Codex, OpenCode, etc.) working in separate tmux panes through a structured task-driven workflow. Instead of ad-hoc communication, agents exchange structured messages with task labels and report tags, creating a traceable collaboration loop. Additional plugins provide code review, learning, and self-reflection capabilities.
┌─────────────┐ task from ┌─────────────┐
│ Agent A │ ───────────────→ │ Agent B │
│ (Sender) │ │ (Receiver) │
│ │ ←─────────────── │ │
└─────────────┘ report from └─────────────┘
Register this repository as a plugin marketplace, then install:
/plugin marketplace add fingergohappy/ai-kit
Install the plugins you need:
/plugin install agentflow@ai-kit
/plugin install tmux@ai-kit
/plugin install git@ai-kit
/plugin install code-kit@ai-kit
/plugin install learning@ai-kit
/plugin install self-learn@ai-kit
After installation, restart Claude Code. Skills will be available with the plugin prefix:
/agentflow:task login-system
/agentflow:dispatch %7 docs/tasks/login_feature.md
/code-kit:evaluate "use postgres vs mysql"
/learning:learn rust lifetimes
/git:commit
/tmux:tmux-send %7 "hello"
claude --plugin-dir /path/to/ai-kit
This repository includes Codex plugin manifests for agentflow, tmux, and git.
Register this repository as a Codex plugin marketplace:
codex plugin marketplace add https://github.com/fingergohappy/ai-kit
Quick AI-driven install/update:
Paste this URL into Codex and ask it to follow the runbook:
follow this install https://github.com/fingergohappy/ai-kit/blob/main/docs/codex_plugin_install_update.md
Detailed guide:
Agent collaboration loop: task → dispatch → evaluate → report → review → redo.
| Skill | Purpose |
|---|---|
agentflow:task | Generate structured task documents (feature / change / task) |
agentflow:dispatch | Send task or fix instructions to a tmux pane for execution |
agentflow:gate-evaluate | Receiver-side input guard — evaluate incoming tasks before execution |
agentflow:report | Report execution results back to the sender |
agentflow:gate-review | Sender-side output guard — review work results, decide pass or redo |
Code review and evaluation utilities for single-agent workflows.
| Skill | Purpose |
|---|---|
code-kit:evaluate | Rigorous evidence-based evaluation (tech selection, architecture, claim verification) |
code-kit:review-init | Analyze project tech stack and generate customized review skills |
code-kit:review-report | Generate structured review report from audit results |
code-kit:fix-review | Fix a specific issue from a review report and update its status |
code-kit:fix-review-all | Batch-fix all pending issues from a review report in parallel |
code-kit:nvim-lsp-init | Generate Neovim LSP environment setup scripts for the project |
Personal learning and note-taking aids.
| Skill | Purpose |
|---|---|
learning:learn | Explain concepts with minimal examples, code analogies, and simplification |
learning:take-note | Generate structured learning notes with runnable code examples |
AI self-reflection and lesson capture.
| Skill | Purpose |
|---|---|
self-learn:learn-from-mistake | After AI is corrected, propose solidifying the lesson as a guardrail rule |
Tmux infrastructure utilities for inter-pane communication and long-running service management.
| Skill | Purpose |
|---|---|
tmux:tmux-send | Send text content to a tmux pane |
tmux:agent-tmux | Start/restart/stop long-running commands in shared tmux session (auto-isolates by project/branch) |
Git worktree and branching utilities.
| Skill | Purpose |
|---|---|
git:rebase-to-root | Rebase worktree feature branch back to root's current branch |
git:commit | Create atomic git commits with validation and conventional commit messages |
/agentflow:task <task-name>
Enter design discussion mode — discuss without writing code, generate document when ready. Outputs to docs/tasks/.
/agentflow:dispatch [loop] <pane_id> <doc-path>
npx claudepluginhub fingergohappy/ai-kit --plugin code-kitAgent collaboration flow: task → dispatch → evaluate → report → review → redo
Tmux utilities for sending commands and text across panes
Personal learning note generation and management skills
fingergohappyAI self-reflection and learning skills for capturing lessons from mistakes
Git worktree and branching utilities for rebasing and atomic commits
Tools to maintain and improve CLAUDE.md files - audit quality, capture session learnings, and keep project memory current.
anthropicsHarness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
affaan-mComprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
JeffallanBrowser automation and end-to-end testing MCP server by Microsoft. Enables Claude to interact with web pages, take screenshots, fill forms, click elements, and perform automated browser testing workflows.
anthropicsMulti-model consensus engine integrating OpenAI Codex CLI, Gemini CLI, and Claude CLI for collaborative code review and problem-solving.
xrf9268-hueComprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review
anthropics