# CTRL_ALT_DESIGN

- **Event:** [Built with Opus 4.7: a Claude Code hackathon](https://cerebralvalley.ai/e/built-with-4-7-hackathon)
- **When:** Apr 21 at 12:00 PM – Apr 27 at 2:00 AM (EDT)
- **Where:** Online
- **Team:** [Elleta McDaniel](https://cerebralvalley.ai/u/CTRL_ALT_DESIGN)
- **GitHub:** https://github.com/emcdanie/chip
- **Demo video:** https://www.loom.com/share/ab0ac2410e3947b49efb5b8166dcd63a
- **Gallery:** https://cerebralvalley.ai/e/built-with-4-7-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/built-with-4-7-hackathon/hackathon/gallery/165

I ship and manage design systems for clients. BELLA is the synthetic stand-in I built for this demo — I can't show real client data, but the workflow is identical.

Every system I onboard has context locked in team members' heads. Every morning I check ten tools by hand — Storybook, Figma, Zeroheight, three platform repos, Jira, Slack, Notion — to find where parity drifted overnight. The decisions get made on gut, because the data's too scattered to use.

Most "design system tools" are static doc sites. They describe the system; nobody watches it.

CHIP is different. It's a control room where an agent watches the system and proposes the diff before the parity drifts. When drift is detected, CHIP files an audit, drafts a roadmap, and waits. I review. I approve. CHIP files the Jira tickets, notifies the platform leads, and writes the diff to the audit ribbon. Every action traceable. No silent merges.

Behind the Ask CHIP input is a Claude Managed Agent on Opus 4.7 with the Notion MCP wired in. When I ask "what does Vitaly say about dashboards", it queries my actual research library — 47 sources from six primary voices on AI-ready design systems.

The dashboard runs the loop on itself. The loop closed.

---

Markdown version of https://cerebralvalley.ai/e/built-with-4-7-hackathon/hackathon/gallery/165. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt