By george11642
Fully autonomous overnight coding agent. Give it a mission, walk away for hours, come back to a better codebase. Supports mission-driven development, codebase improvement, plan execution, and deep research modes. Fresh-context loop pattern with file-based memory for unlimited runtime.
Verify service credentials and auth state for autonomous operations
Generate a comprehensive report of what autopilot accomplished.
Resume a stopped autopilot session from its handoff state.
Check autonomous loop progress, completed tasks, and current state.
Gracefully stop the autonomous loop.
Manages safe production deployments via Vercel canary releases. Deploys to preview, monitors for errors, then promotes or reports rollback.
Independent post-milestone code quality evaluator. Reviews all changes for quality, security, and test coverage after a milestone or phase completes. Read-only — never modifies code.
Code implementation agent for autopilot. Writes, edits, and tests code changes. Follows existing patterns, handles edge cases, and self-reviews before reporting completion.
Codebase health analysis agent for autopilot improve mode. Scans for test gaps, lint issues, performance problems, security risks, and code quality issues. Outputs a prioritized improvement plan.
Generates legal documents (Terms of Service, Privacy Policy, Cookie Policy, AUP) from app metadata. Outputs ready-to-customize templates.
Auto-triggers when running inside an autopilot loop iteration. Provides the core methodology for autonomous task execution: orient from file state, implement one task, verify, commit, update state. Activates on: autopilot iteration, autonomous mode, overnight coding, unattended execution.
Auto-triggers when context is getting large or when running long sessions. Provides smart context management: when to write handoff files, how to structure state for session chaining, how to spawn fresh sessions. Activates on: long session, context pressure, session handoff, memory management, running out of context.
Uses power tools
Uses Bash, Write, or Edit tools
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.
A production-grade orchestration framework for Claude Code with 36 domain skills, 22 hooks, 2-mode autonomy system, RTK token optimization, and autopilot supervisor.
+-----------------+
| CLAUDE.md | Global orchestrator rules
| (always loaded) | Delegation, dispatch, verification
+-----------------+
|
+------------+------------+
| |
+--------v--------+ +--------v--------+
| behavior.json | | settings.json |
| autonomy: auto | | hooks, plugins, |
| toggles: r/o | | MCP, env vars |
+-----------------+ +-----------------+
| |
+---------+---------+ +---------+---------+
| | | |
+---v---+ +----v----+ +----v----+
| Modes | | Hooks | | Skills |
| 2-mode| | 22 active| | 36 total|
+-------+ +----------+ +---------+
| |
+---------+---------+ +----------+----------+
| | | | | | | | | |
boot router gate RTK ... L1 L2 L3 ...
Controlled via behavior.json:
| Mode | Behavior |
|---|---|
| autonomous | Never ask, never checkpoint. Proceed on everything. Only escalate payments. |
| guided | Checkpoint before destructive/irreversible actions. Confirm at major milestones. Still autonomous on non-destructive work. |
Switch modes: /mode guided, /mode autonomous, /mode readonly (toggle)
| Tier | Count | Purpose | Examples |
|---|---|---|---|
| L1 Meta | 2 | Framework control | mode, router |
| L2 Domain | 13 | Full-stack dev domains | web-frontend, backend-data, python-dev |
| L3 Atomic | 15+ | Single-purpose tools | deploy-convex, git-workflow, mcp-gemini |
| Utility | 6 | Framework utilities | completion-council, consolidate-memory, dashboard |
Hooks fire at specific lifecycle events. See MANIFEST.md for the complete catalog.
| Event | Hooks |
|---|---|
| SessionStart | boot.js, gsd-check-update.js |
| UserPromptSubmit | router.js, gate-enforcer.js |
| PreToolUse | dangerous-cmd-block.js, rtk-rewrite.sh, readonly-guard.js, convention-check.js |
| PostToolUse | context-tracker.js, self-heal.js, mid-session-skill-loader.js, budget-breaker.js, sycophancy-check.js |
| Stop | quality-gate.js, screenshot-verify.js |
| PreCompact | precompact-handoff.js |
| StatusLine | statusline.js |
Transparent CLI proxy that rewrites commands (e.g., git status -> rtk git status) for 60-90% token savings. The rtk-rewrite.sh hook handles this automatically.
curl -fsSL https://raw.githubusercontent.com/george11642/george-plugins/main/setup.sh | bash
/plugin marketplace add george11642/george-plugins
/plugin install george-setup@george-plugins
/george-setup:install
# Browser automation (third-party sites, personal auth)
/plugin marketplace add obra/superpowers-marketplace
/plugin install superpowers-chrome@superpowers-marketplace
# GSD milestone execution
/plugin install ralph-gsd@george-plugins
# LSP support
/plugin marketplace add anthropics/claude-plugins-official
/plugin install typescript-lsp@claude-plugins-official
/plugin install pyright-lsp@claude-plugins-official
The install command copies templates/settings-template.json to ~/.claude/settings.json. Review and adjust paths.
Copy templates/behavior.json to ~/.claude/behavior.json:
{
"autonomy": "autonomous",
"toggles": { "readonly": false }
}
| Plugin | Status | Description |
|---|---|---|
george-setup | Core | Skills, hooks, modes, CLAUDE.md, RTK |
ralph-gsd | Enabled | ROADMAP-driven milestone execution orchestrator |
autopilot | Available | Autonomous overnight coding agent (systemd supervisor) |
ralph-marketer-v2 | Available | Marketing automation agent |
| Plugin | Source |
|---|---|
superpowers-chrome | superpowers-marketplace |
ralph-gsd | george-plugins |
typescript-lsp | claude-plugins-official |
pyright-lsp | claude-plugins-official |
npx claudepluginhub george11642/george-plugins --plugin autopilotAutonomous milestone execution orchestrator with parallel phase execution, dependency graph parsing, and fully autonomous DEFERRED.md item resolution (security fixes, outreach, test failures) — no human involvement required.
Complete Claude Code power-user setup: T0-T4 orchestration engine, context-aware hooks, MCP server sources, and a one-command installer. Replicate George's exact Claude Code workflow.
Claude Code settings and skills for spec-driven development workflows
Autonomous agent orchestrator for full development lifecycles with zero human input, session budget management, and crash recovery
Harness for Claude Code — skills, /harness:* slash commands, persona subagents, lifecycle hooks, and MCP tools without per-repo `harness setup`. Sibling plugins exist for Cursor, Gemini CLI, and Codex.
Autonomous session orchestrator for Claude Code - manages multi-phase development workflows
Autonomous multi-agent development framework with spec-driven sprints and convergent iteration
Automation bootstrap for Claude Code. Analyzes any project and generates a complete .claude/ structure with agents, pipelines, skills, memory, hooks, and settings.