# AI use cases for Overclick (ustoppble/overclick) Repo: [ustoppble/overclick](https://github.com/ustoppble/overclick) | Use cases: 4 | Page: https://dedupedai.com/use-cases/ustoppble-overclick/ OverClick is a self-hosted, open source task board where humans write task contracts and AI coding agents claim, do, and report back on the work over MCP. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Run a private task board on your own server: https://dedupedai.com/use-cases/ustoppble-overclick/self-host-private-task-board/ - Let Claude Code claim and deliver board cards: https://dedupedai.com/use-cases/ustoppble-overclick/connect-claude-code-to-board/ - Install the Claude Code plugin for the full workflow: https://dedupedai.com/use-cases/ustoppble-overclick/install-claude-code-plugin-workflow/ - Review agent output against your own test scripts: https://dedupedai.com/use-cases/ustoppble-overclick/validate-agent-work-against-test-scripts/ ## How to run a private task board on your own server with Overclick Repo: [ustoppble/overclick](https://github.com/ustoppble/overclick) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Start a self-hosted board where your task data never leaves your own instance. **Who it is for:** A developer or team lead who wants a task board for human and AI-agent work without sending data to a hosted service. **The problem:** Hosted task tools send work data to third parties, and many teams cannot accept that for internal engineering work. **The outcome:** A running OverClick instance on your machine or server, with a local admin account and an onboarding flow for your first project. ### Steps 1. Run `git clone https://github.com/ustoppble/overclick && cd overclick` 2. Run `export AUTH_SECRET="$(openssl rand -base64 32)"` and store the value in your protected environment so restarts reuse it 3. Run `docker compose up --build` 4. Open http://localhost:3000 and create the local admin account with an e-mail and password 5. Follow the 3-step onboarding: project, executors, connect your agent ### Prompt to paste into your agent ```text Clone the ustoppble/overclick repository, generate AUTH_SECRET with openssl, and start the stack with docker compose up --build. Confirm the web app responds at http://localhost:3000 and report the URL and the step where it stopped, if any. ``` ### Caveat The README says the onboarding wizard, settings, and insights are still landing, so expect a v0.1 setup with some flows not yet built. Risk, credentials: AUTH_SECRET signs logins and must be unique per installation, so store the generated value safely and do not commit it to git. 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/ustoppble-overclick/self-host-private-task-board/ ## How to let Claude Code claim and deliver board cards with Overclick Repo: [ustoppble/overclick](https://github.com/ustoppble/overclick) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Connect Claude Code over MCP so it picks up cards, works them, and hands back evidence. **Who it is for:** A developer with a running OverClick board who wants a coding agent to take the next task from the backlog without manual setup each time. **The problem:** Assigning work to a coding agent by copying task text into a chat loses the contract, the evidence, and the record of what the run cost. **The outcome:** Claude Code is connected to your board, claims a card, does the work on your machine, and delivers a handoff with a summary and telemetry. ### Steps 1. Create an agent token for your board in the OverClick admin interface 2. Run `claude mcp add --transport http overclick http://:3000/mcp --header "Authorization: Bearer "` (inferred, check the README) 3. In your terminal, tell Claude Code: "grab the next task from the board" 4. Review the handoff card with its summary, evidence, branch or PR links, and tokens per model 5. Open the card, run its test script, and mark it Validated if it passes, or reopen it with a comment ### Prompt to paste into your agent ```text Using the ustoppble/overclick MCP server at /mcp, claim the next open task and deliver it with a summary, evidence, and telemetry. Report the card ID, the branch name, and the token totals per model, and confirm the card moved to In progress before you started work. ``` ### Caveat The README's example command has placeholder spacing that you must fill in yourself, and the configuration tools stay off unless a per-token manage flag is enabled. Risk, credentials: The bearer token grants agent access to your board, so keep it out of shared shell history and rotate it if it leaks. 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/ustoppble-overclick/connect-claude-code-to-board/ ## How to install the Claude Code plugin for the full workflow with Overclick Repo: [ustoppble/overclick](https://github.com/ustoppble/overclick) | Difficulty: Beginner | Time: 15 minutes | Basis: From the README Add the plugin for OVERCLICK.md, the /overclick commands, and lifecycle hooks. **Who it is for:** A Claude Code user who already connected the MCP server and wants the standard OverClick commands and hooks in every session. **The problem:** The MCP connection alone does not give the canonical workflow file, slash commands, or lifecycle hooks, so each session needs manual setup. **The outcome:** The Overclick plugin is installed, verified as enabled, and available for updates from the repository's marketplace. ### Steps 1. Run `claude plugin marketplace add ustoppble/overclick` 2. Run `claude plugin install overclick@overclick` 3. Run `claude plugin list` and confirm Overclick@Overclick is enabled with a version 4. Run `claude plugin details overclick` and confirm the components are listed 5. Run `claude plugin update overclick@overclick` when a later release is available ### Prompt to paste into your agent ```text Add the ustoppble/overclick plugin marketplace to Claude Code, install overclick@overclick, and verify it with claude plugin list and claude plugin details overclick. Report the installed version and the listed components, and do not treat the install message alone as proof. ``` ### Caveat The README says other CLIs need install.sh instead, and that path is documented in docs/design/plugin.md rather than shown here. 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/ustoppble-overclick/install-claude-code-plugin-workflow/ ## How to review agent output against your own test scripts with Overclick Repo: [ustoppble/overclick](https://github.com/ustoppble/overclick) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Accept or reopen agent deliveries using the test script written on each card. **Who it is for:** A reviewer who delegates engineering tasks to agents and needs a consistent way to decide whether the work is done. **The problem:** Agent output is often judged by a quick look, so merged code is mistaken for finished work and nothing records what the run cost. **The outcome:** Each card is marked Validated only after a human runs the test script, and reopened cards show the reviewer's comment on the agent's next claim. ### Steps 1. Create a card with What, Why, and How to confirm, where How to confirm is a plain-language test script 2. Let your agent claim and deliver the card through the board 3. Read the handoff with its summary, evidence, branch or PR links, and tokens and time per model 4. Run the test script yourself against the delivered branch 5. Stamp Validated if the script passes, or reopen the card with a comment so the agent sees it on the next claim ### Prompt to paste into your agent ```text Pick the delivered card from the ustoppble/overclick board, run its How to confirm script against the branch named in the handoff, and report pass or fail with the evidence. Do not mark it Validated, since only a human stamps that status. ``` ### Caveat The board does not run your test script for you, so the review still depends on a human or an agent executing it correctly. 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/ustoppble-overclick/validate-agent-work-against-test-scripts/