https://github.com/schoolai/shipyard
Verify AI agent work with collaborative review and proof-of-work artifacts
https://github.com/schoolai/shipyard
ai-agents claude-code code-review collaboration mcp proof-of-work task-management typescript
Last synced: 1 day ago
JSON representation
Verify AI agent work with collaborative review and proof-of-work artifacts
- Host: GitHub
- URL: https://github.com/schoolai/shipyard
- Owner: SchoolAI
- License: other
- Created: 2026-01-03T05:10:31.000Z (about 2 months ago)
- Default Branch: main
- Last Pushed: 2026-02-19T08:24:36.000Z (4 days ago)
- Last Synced: 2026-02-19T08:56:47.095Z (4 days ago)
- Topics: ai-agents, claude-code, code-review, collaboration, mcp, proof-of-work, task-management, typescript
- Language: TypeScript
- Size: 11.4 MB
- Stars: 1
- Watchers: 0
- Forks: 1
- Open Issues: 11
-
Metadata Files:
- Readme: README.md
- License: LICENSE.md
- Agents: AGENTS.md
Awesome Lists containing this project
README
Shipyard
Ship responsibly.
Agent management hub for human-agent collaboration.
---
## The Problem
You're managing multiple AI agents (Claude, Cursor, Devin), but there's no workspace where humans and agents collaborate together:
- **No verification** — Agent says "done" but you have no proof
- **No collaboration layer** — Humans review in GitHub, agents work in chat logs
- **No feedback loop** — You approve work, but the agent never sees it
Shipyard is the collaboration workspace for mixed human-agent teams. Agents create tasks with proof. Humans review in real-time. Feedback flows both ways.
## How It Works
```
┌──────────────────────────┐
│ Session Server │
│ (CF Workers + Durable │
│ Objects) │
│ • Auth (Shipyard JWT) │
│ • WebRTC signaling │
└─────────┬──────────────────┘
│
┌──────────────────┼──────────────────┐
│ │ │
┌────────▼───────┐ ┌──────▼───────┐ ┌───────▼───────┐
│ Developer A │ │ Browser B │ │ Phone C │
│ Machine │ │ (reviewer) │ │ (mobile) │
│ │ │ │ │ │
│ ┌───────────┐ │ │ IndexedDB │ │ IndexedDB │
│ │ Daemon/MCP │ │ │ persistence │ │ persistence │
│ │ Server │ │ │ │ │ │
│ └─────┬──────┘ │ └───────────────┘ └───────────────┘
│ │ MCP │
│ ┌────▼──────┐ │ All peers sync via
│ │ AI Agents │ │ Loro CRDT over WebRTC
│ └────────────┘ │
└──────────────────┘
```
1. **Agent creates task** via MCP → Browser opens automatically
2. **Reviewers join** via shared URL → Real-time P2P sync
3. **Add comments**, approve, or request changes → Agent sees feedback
4. **Agent uploads artifacts** (screenshots, videos) as proof of work
5. **Task auto-completes** when all deliverables have receipts
## Quick Start
```bash
# Install the daemon
npm install -g @schoolai/shipyard
# Log in with GitHub (opens browser for device auth)
shipyard login
# Verify you're authenticated
shipyard login --check
# Start the daemon (connects to signaling, syncs via CRDT, runs agents)
ANTHROPIC_API_KEY=sk-ant-... shipyard-daemon --serve
```
See the [Installation Guide](./docs/installation.md) for editor-specific MCP setup (Claude Code, Cursor, VS Code, etc.).
## Current State
Shipyard is in active development. The MCP server and web app are being rebuilt on a clean foundation:
| Component | Description |
|-----------|-------------|
| [**Session Server**](./apps/session-server) | Auth + WebRTC signaling (Cloudflare Workers + Durable Objects) |
| [**OG Proxy**](./apps/og-proxy-worker) | Open Graph meta tags for social link previews |
| [**Loro Schema**](./packages/loro-schema) | CRDT Shape definitions, typed documents, helpers |
| [**Daemon**](./apps/daemon) | CLI agent runner + device flow auth (`@schoolai/shipyard` on npm) |
| [**Session**](./packages/session) | Session/auth shared types and client |
**Tech stack:** Loro CRDT (loro-extended), TipTap editor, HeroUI v3, Tailwind v4, Cloudflare Workers
## Data & Privacy
| Data | Where It Lives | Control |
|------|---------------|---------|
| Task content | Browser (IndexedDB) + P2P sync | You own it |
| Artifacts | Optional: GitHub (your repo, orphan branch) | You own it |
| MCP server | Runs locally | Never leaves your machine |
| URLs | Encoded snapshots | Shareable, regenerable |
**No telemetry. No cloud storage. GitHub optional.**
## Documentation
| Doc | Description |
|-----|-------------|
| **[Development](./docs/development.md)** | Local setup, running services |
| **[Architecture](./docs/architecture.md)** | Data model, sync topology, tech choices |
| **[Engineering Standards](./docs/engineering-standards.md)** | Code quality, testing philosophy |
| **[Installation](./docs/installation.md)** | Installing Shipyard across editors and CLI |
## Contributing
We value **ideas over implementations**. Please start with discussion:
1. **Bug reports** — [Open an issue](https://github.com/SchoolAI/shipyard/issues/new)
2. **Feature ideas** — [Start a discussion](https://github.com/SchoolAI/shipyard/discussions/new)
3. **Questions** — [Ask in discussions](https://github.com/SchoolAI/shipyard/discussions)
PRs without a linked, approved issue may be closed. AI-assisted contributions are welcome — what matters is that **you** understand what you're submitting.
## License
[FSL-1.1-ALv2](./LICENSE.md) (Functional Source License)
- **Free** for all non-competing use
- **Converts to Apache 2.0** automatically in 2 years