# Gatz

- **Event:** [Built with Opus 4.6: a Claude Code hackathon](https://cerebralvalley.ai/e/claude-code-hackathon)
- **When:** Feb 10 at 12:00 PM – Feb 17 at 10:00 AM (EST)
- **Where:** Location TBA
- **Team:** [Ramzi Lb](https://cerebralvalley.ai/u/rmzlb)
- **GitHub:** https://github.com/rmzlb/baaton
- **Demo video:** https://youtu.be/usX1k144okA
- **Gallery:** https://cerebralvalley.ai/e/claude-code-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/claude-code-hackathon/hackathon/gallery/55

Most AI dev tools try to own the whole pipeline — prompt, generate, deploy, done. I don't buy it. When your project gets real, all-in-one tools fall apart. Baaton takes the opposite bet: your agent is already good at coding. What's missing is the layer above. An orchestration board where AI agents (Claude Code, Codex) are first-class. They connect via REST API, pull tickets, update status, and post structured summaries — files changed, test results, PR links. The human stays in control. - Kanban with drag-and-drop, bulk actions, shortcuts - Agent API with key-based auth - TLDR system — agents post structured debriefs after work - Public forms with LLM auto-qualification for bugs - Multi-tenant via Clerk orgs with Postgres RLS - Cross-org dashboard across all your teams. Stack: React 19, Vite 8, Rust/Axum, Supabase Postgres, Clerk, Railway. ~30 endpoints with compile-time checked SQL. The future isn't one magic tool. It's specialized agents on a shared board. Baaton is that board.

---

Markdown version of https://cerebralvalley.ai/e/claude-code-hackathon/hackathon/gallery/55. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt