Coding & Developmentβ¨geminiπ€claudeβ¨οΈcursorπwindsurfπuniversal
Autonomous Dev Team
Turns GitHub issues into merged PRs with zero human intervention. Powered by OpenClaw, supports Claude Code, Codex CLI, and Kiro CLI.
Quick Start
Install this Skill
npx skills add zxkane/autonomous-dev-team# Autonomous Dev Team
A fully automated development pipeline that turns GitHub issues into merged pull requests β no human intervention required. It scans for issues labeled `autonomous`, dispatches a **Dev Agent** to implement the feature with tests in an isolated worktree, and hands off to a **Review Agent** for code review with optional E2E verification. The entire cycle runs unattended on a cron schedule.
Supports multiple coding agent CLIs β Claude Code, Codex CLI, Kiro CLI, opencode, Cursor Agent, Gemini CLI, and most CLIs with a `-p <prompt>` non-interactive flag β via a pluggable agent abstraction layer.
## Getting Started
### Option A: Install as Portable Skills (Recommended)
Install the skills into **any** of 40+ supported coding agents with a single command:
```bash
npx skills add zxkane/autonomous-dev-team
```
This installs the following skills into your agent:
| Skill | Description |
|-------|-------------|
| **autonomous-dev** | TDD workflow with git worktree isolation, design canvas, test-first development, code review, and CI verification |
| **autonomous-review** | PR code review with checklist verification, merge conflict resolution, E2E testing, and auto-merge |
| **autonomous-dispatcher** | GitHub issue scanner that dispatches dev and review agents on a cron schedule |
| **create-issue** | Structured GitHub issue creation with templates, autonomous label guidance, and workspace change attachment |
Supported agents include Claude Code, Cursor, Windsurf, Gemini CLI, Kiro CLI, and [many more](https://skills.sh). See the [skills.sh docs](https://skills.sh/docs) for the full list and usage guide.
#### Post-Install: Enable Workflow Hooks (Claude Code / Kiro CLI)
The `autonomous-dev` and `autonomous-review` skills define Claude Code hooks in their SKILL.md frontmatter. These hooks enforce the TDD workflow (block direct pushes to main, require code review before commit, etc.). The hook commands reference `$CLAUDE_PROJECT_DIR/hooks/...`, but `npx skills add` places hook scripts inside `.claude/skills/autonomous-common/hooks/`. You need to create symlinks so the paths resolve:
```bash
# From your project root β create symlinks after npx skills add:
ln -sf .claude/skills/autonomous-common/hooks hooks
ln -sf .claude/skills/autonomous-dispatcher/scripts scripts
```
> **Why symlinks?** `npx skills add` copies each skill directory into `.claude/skills/`, but hook commands use `$CLAUDE_PROJECT_DIR/hooks/` (the project root). The symlinks bridge this gap. The `scripts/` symlink enables agent-callable utility scripts referenced by the skills (e.g., `scripts/gh-as-user.sh`).
**Required Claude Code plugins** (add to `.claude/settings.json` under `enabledPlugins`):
```json
{
"enabledPlugins": {
"code-simplifier@claude-plugins-official": true,
"pr-review-toolkit@claude-plugins-official": true
}
}
```
> **Note:** If your IDE does not support hooks (Cursor, Windsurf, Gemini CLI), the skills still work β follow each workflow step manually. Hooks only provide automatic enforcement.
### Option B: Use as GitHub Template (Full Pipeline)
For the complete autonomous pipeline β including hooks, wrapper scripts, dispatcher cron, and GitHub App auth:
1. **Clone and configure**:
```bash
gh repo create my-project --template zxkane/autonomous-dev-team
cd my-project
cp scripts/autonomous.conf.example scripts/autonomous.conf
# Edit autonomous.conf β set REPO, PROJECT_DIR, agent CLI, etc.
```
2. **Set up GitHub labels**:
```bash
bash scripts/setup-labels.sh owner/repo
```
3. **Install a dispatcher orchestration host and schedule the tick.** The dispatcher skill is a standard skills.sh skill β anything that can run a coding-agent CLI on a schedule works. Pick one:
| Host | When to use | Example tick command |
|------|------------|---------------------|
| [OpenClaw](https://github.com/OpenClaw/OpenClaw) (recommended) | Purpose-built skill runtime; first-class support for skill invocation. | `openclaw run skills/autonomous-dispatcher/SKILL.md` |
| Plain cron + agent CLI | Zero extra infra; you already have an agent CLI installed. | `bash skills/autonomous-dispatcher/scripts/dispatcher-tick.sh` (the tick script doesn't need an agent CLI to run β see note below). |
| Claude Cowork / GitHub Actions schedule / any scheduled-agent runtime | You want the dispatcher to run on managed infra rather than a local box. | Schedule a job that runs `bash skills/autonomous-dispatcher/scripts/dispatcher-tick.sh` (or `dispatcher-multi-tick.sh` for multi-project). |
```cron
# Example A β OpenClaw via OS cron, every 5 minutes
*/5 * * * * cd /path/to/project && openclaw run skills/autonomous-dispatcher/SKILL.md
# Example B β plain cron without OpenClaw
*/5 * * * * cd /path/to/project && bash skills/autonomous-dispatcher/scripts/dispatcher-tick.sh
```
The tick script (`dispatcher-tick.sh`) is host-agnostic β it only needs `gh`, `jq`, and a per-project `
... (truncated)Repository
21stars
5forks
Details
Related Skills
More coding & development skills you might like
Reticle
Coding & DevelopmentAI agents can generate code, but still struggle to understand what they build. Reticle brings Jev-style machine-native runtime perception to web & desktop applications.
π€claudeπuniversal
1.2k
190
View
agentscope-java
Coding & DevelopmentExpert Java developer skill for AgentScope Java framework - a reactive, message-driven multi-agent system built on Project Reactor. Use when working with reactive programming, LLM integration, agent orchestration, multi-agent systems, or when the user mentions AgentScope, ReActAgent, Mono/Flux, Project Reactor, or Java agent development. Specializes in non-blocking code, tool integration, hooks, pipelines, and production-ready agent applications.
πuniversal
6.0k
1.5k
View
ospec
Coding & DevelopmentDocument-driven OSpec workflow for initialization, change/goal routing, validation, archiving, and durable project knowledge.
β¨geminiπ€claudeπ¬gptπuniversal
524
34
View
misakanet-failure-memory
Coding & Developmentπ A zero-dependency, git-backed micro-lesson library for AI Agents to asynchronously share and search verified debugging experience. Python stdlib only. | https://misakanet.org
π€claudeβ¨οΈcursorπuniversal
424
158
View