autonomous-builder
Full-stack software development agent for design, implementation, testing, and deployment. Use when the user explicitly asks for end-to-end project creation, feature development, bug fixing, or code refactoring.
npx skills add majiayu000/claude-skill-registry --skill autonomous-builder-foryourhealth111-pix-vibe-skills --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
A fully autonomous software development agent that handles end-to-end lifecycle tasks: requirements analysis, architecture design, implementation, testing, debugging, and deployment. It operates in sessions with an INITIALIZER AGENT that sets up project structure, tests, and Git state, and subsequent BUILDER AGENTS that carry out feature implementation, browser-based UI testing, and iterative progress updates. Progress is stored in file-based state (feature_list.json and related .builder files), and progress is committed to Git as atomic units. It automatically continues to the next feature after completing one, with a 3-second delay between sessions.
How it works
- It uses a two-agent model: Initializer Agent reads requirements, creates project structure, generates feature_list.json, initializes a Git repo, optionally captures a GitHub URL, creates README.md and PLANNING.md, commits the initial state, and may push to GitHub and open issues.
- Builder Agents run in fresh context per session and perform steps: get context (pwd, ls, git log, progress), start/verify server, verify regression tests, pick the next non-passing feature, implement that feature, perform a browser automation test, update feature_list.json, generate a workflow report, commit and push to GitHub, and update progress notes, then auto-exit (with a 3-second delay) to continue with the next feature if any remain.
- The design emphasizes file-based state persistence (via feature_list.json and .builder directory), Git commits as the state anchor, and automated progression without user intervention beyond initial requirements.
When to use it
Use when the user explicitly requests end-to-end project creation, feature development, bug fixing, or code refactoring that spans multiple steps and requires full lifecycle ownership. It’s intended for starting new projects from full specs, continuing a previously initialized .builder project, driving broad feature builds across steps, or performing explicit refactors across the codebase.
What it can touch
- Tools declared in frontmatter include: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Skill, Task, ToolSearch, mcp__ide__executeCode, mcp__ide__getDiagnostics. The skill uses these to perform file operations, code execution, web lookups, and UI/browser testing via MCP actions. It relies on a file-based state under .builder/ and standard Git operations for versioning.
Caveats
- Requires explicit user approval for operations that could affect system stability, data integrity, or files outside the workspace.
- Security-critical systems and production deployments require user confirmation.
- The workflow assumes a non-interactive, unattended mode with auto-continue behavior between features; human intervention may be needed for complex decisions or environment changes.
# Autonomous Builder A fully autonomous software development agent that handles the complete software lifecycle: requirements analysis, architecture design, implementation, testing, debugging, and deployment. ## Architecture Pattern: Two-Agent Model **Based on Anthropic's official claude-quickstarts architecture** ``` ┌─────────────────────────────────────────────────────────────────┐ │ TWO-AGENT ARCHITECTURE │ ├─────────────────────────────────────────────────────────────────┤ │ │ │ SESSION 1: INITIALIZER AGENT │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ • Read requirements / spec │ │ │ │ • Create project structure │ │ │ │ • Generate feature_list.json (200+ tests) │ │ │ │ • Initialize Git repository │ │ │ │ • ✨ Prompt for GitHub URL (optional) │ │ │ │ • ✨ Create README.md & PLANNING.md │ │ │ │ • Commit initial state │ │ │ │ • ✨ Push to GitHub & create issues │ │ │ └─────────────────────────────────────────────────────────┘ │ │ │ │ │ feature_list.json │ │ (Single Source of Truth) │ │ │ │ │ SESSIONS 2+: BUILDER AGENT (fresh context each session) │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ Step 1: Get Context (pwd, ls, git log, prog
- Architecture Pattern: Two-Agent Model
- Core Philosophy
- When to Use This Skill
- Not For / Boundaries
- Quick Reference
- Session Continuity (Auto-Resume)
- Automatic Task Queue
- Directory Structure
- Skill Recommendations & Router Handoff
- MCP Auto-Integration & Human-like Computer Control
- Auto-Restart & Self-Supervision
- Core Workflow Phases
- State File Schema
- 3-Strike Error Recovery
Self-supervision loop while true; do if [ -f "$STATE_FILE" ]; then if [ "$STATUS" = "completed" ]; then echo "[$(date)] Project completed. Exiting supervisor." exit 0 fi if [ -n "$LAST_ACTIVITY" ]; then echo "[$(date)] Starting Claude session..." claude --skill autonomous-builder --project "$PROJECT_DIR" --dangerously-skip-permissions
What does the autonomous-builder skill do?
Full-stack software development agent for design, implementation, testing, and deployment. Use when the user explicitly asks for end-to-end project creation, feature development, bug fixing, or code refactoring.
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill autonomous-builder-foryourhealth111-pix-vibe-skills --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.