# AI use cases for Doop (kgoedecke/doop) Repo: [kgoedecke/doop](https://github.com/kgoedecke/doop) | Use cases: 4 | Page: https://dedupedai.com/use-cases/kgoedecke-doop/ Doop is an open-source multiplayer design canvas where humans and AI agents design together live, with a built-in MCP server. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Design a page section live with Claude Code on a canvas: https://dedupedai.com/use-cases/kgoedecke-doop/live-section-design-with-claude-code/ - Self-host a private team design canvas with Docker: https://dedupedai.com/use-cases/kgoedecke-doop/self-host-team-canvas-with-docker/ - Queue design cards for the built-in Doop Agent: https://dedupedai.com/use-cases/kgoedecke-doop/queue-design-cards-for-doop-agent/ - Turn pinned frames into durable style rules for agents: https://dedupedai.com/use-cases/kgoedecke-doop/build-style-rules-from-pinned-frames/ ## How to design a page section live with Claude Code on a canvas with Doop Repo: [kgoedecke/doop](https://github.com/kgoedecke/doop) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Connect Claude Code over MCP and watch it stream a frame onto your canvas as you follow along. **Who it is for:** A designer or developer who wants an AI agent to sketch a section, such as pricing, while watching it appear next to their cursor. **The problem:** Prompting an AI for HTML and refreshing a browser to check each change is slow and disconnected from the design. **The outcome:** A new frame with a streamed section on your canvas, created by Claude Code acting as you, visible in the activity feed. ### Steps 1. Run `git clone https://github.com/kgoedecke/doop && cd doop` 2. Run `bun install` 3. Run `bun run dev` and open http://localhost:4300 4. Create or open a canvas, then run `claude mcp add --transport http doop http://localhost:4300/mcp` 5. Approve the OAuth prompt in the browser, then ask Claude Code to design a pricing section on your canvas ### Prompt to paste into your agent ```text In the repo kgoedecke/doop, the Doop server is running on http://localhost:4300 with the MCP endpoint at /mcp. Using the doop MCP server, create one frame on the canvas with a three-tier pricing section, then call get_canvas to confirm the new frame is listed. Report the frame id and the number of frames on the canvas. ``` ### Caveat The README describes this flow on a local dev server, and the agent works as the approving user, so only connect accounts you trust to edit the canvas. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/kgoedecke-doop/live-section-design-with-claude-code/ ## How to self-host a private team design canvas with Docker with Doop Repo: [kgoedecke/doop](https://github.com/kgoedecke/doop) | Difficulty: Advanced | Time: 45 minutes | Basis: From the README Run Doop's production build with Docker so your team designs on infrastructure you control. **Who it is for:** A team lead or ops person who needs a private shared design space and does not want to use the hosted service. **The problem:** Teams that want shared AI-assisted design often cannot send work to a hosted tool they do not control. **The outcome:** Doop running on port 4400 with app and Postgres in containers, ready for invited collaborators. ### Steps 1. Run `git clone https://github.com/kgoedecke/doop && cd doop` 2. Run `BETTER_AUTH_SECRET=$(openssl rand -hex 32) docker compose up -d` 3. Open http://localhost:4400 in a browser 4. Sign up, create a canvas, and invite collaborators by email 5. Optionally set DATABASE_URL to a real Postgres, as the README describes ### Prompt to paste into your agent ```text In the repo kgoedecke/doop, start the production stack with `BETTER_AUTH_SECRET=$(openssl rand -hex 32) docker compose up -d`. Confirm the app answers on http://localhost:4400 and report the HTTP status code returned for the root URL. ``` ### Caveat The README gives no guidance on backups, TLS or upgrades, so plan those yourself before putting real client work on it. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/kgoedecke-doop/self-host-team-canvas-with-docker/ ## How to queue design cards for the built-in Doop Agent with Doop Repo: [kgoedecke/doop](https://github.com/kgoedecke/doop) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Turn on the built-in agent with an Anthropic key and let it pick up cards without a client. **Who it is for:** A product builder who wants design tasks done in the background, without connecting an MCP client. **The problem:** Design requests sit in a backlog until someone opens a tool and does them by hand. **The outcome:** A queued card or @mention that the Doop Agent claims, designs, and reports back in the canvas chat. ### Steps 1. Run `git clone https://github.com/kgoedecke/doop && cd doop` and then `bun install` 2. Add `ANTHROPIC_API_KEY=sk-ant-...` to a .env file in the repo 3. Run `bun run dev` and check the startup banner for the agent state 4. Open a canvas, queue a board card, or @mention a role on an element comment 5. Watch the working strip and the Agents panel, then review the frame it builds ### Prompt to paste into your agent ```text In the repo kgoedecke/doop, with ANTHROPIC_API_KEY set in .env, start the dev server with `bun run dev`. Queue one board card on canvas asking for a hero section, then check the Agents panel and report whether the task was claimed and which frame it produced. ``` ### Caveat Without ANTHROPIC_API_KEY or a connected account the Doop Agent stays off and queued cards just wait, as the README says. Risk, credentials: The server key pays for free tasks and connected model accounts hold tokens, so keep .env and Settings credentials out of shared repos and screenshots. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/kgoedecke-doop/queue-design-cards-for-doop-agent/ ## How to turn pinned frames into durable style rules for agents with Doop Repo: [kgoedecke/doop](https://github.com/kgoedecke/doop) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Pin exemplar frames and let the guideline distiller propose rules every agent follows. **Who it is for:** A design lead who wants consistent output from every agent working on the same canvas. **The problem:** Each agent or designer drifts in style, and the decisions behind a design are lost. **The outcome:** A set of distiller-proposed style rules, backed by pinned exemplars and captured decisions, that agents read before designing. ### Steps 1. Set `ANTHROPIC_API_KEY` in .env, since the distiller uses that key and turns off without it 2. Run `bun run dev` and open your canvas at http://localhost:4300 3. Pin two or three exemplar frames that show the style you want 4. Capture a few design decisions in the canvas 5. Run the distiller and review the proposed rules before accepting them ### Prompt to paste into your agent ```text In the repo kgoedecke/doop, with ANTHROPIC_API_KEY set, pin three exemplar frames on canvas , then run the guideline distiller. Report the rules it proposes and state whether each one cites a pinned frame or a captured decision. ``` ### Caveat The README does not say how the distiller chooses rules or how to edit them afterward, so review every proposal before treating it as a team standard. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/kgoedecke-doop/build-style-rules-from-pinned-frames/