← All runbooks
gooseworks-ai / capabilities-create-dashboard

create-dashboard

This runbook converts the `create-dashboard` skill into a repeatable agent workflow for building a custom dashboard. The source skill says: Create a custom web dashboard (React + Vite + Express) inside your sandbox to visuali

agent codexmodel gpt-5.5snapshot python312-uveval programmatic7 stepsv1.0.0

Deploy create-dashboard to your jetty.io

One-click installs this runbook into a collection on your Jetty account. You can run it from the Spot dashboard, schedule it, or pipe inputs in via the API.

Deploy on jetty.io →View source
The shape of the run

7 steps · start to finish.

  1. 1
    Step 1

    Environment Setup

    ▶
    1. Create /app/results if it does not already exist.
    2. Confirm /home/user/dashboard exists before running npm commands.
    3. Run npm install only from /home/user/dashboard if dependencies are missing.
    4. Confirm port 3847 is available, or stop the stale process that owns it.
    mkdir -p /app/results
test -d /home/user/dashboard
cd /home/user/dashboard
npm --version
node --version
  2. 2
    Step 2

    Inspect Dashboard Source

    ▶

    Read the existing dashboard files before editing. Treat `/home/user/dashboard` as the runnable project root even when files are symlinked to a workspace path. Identify available data access helpers, API routes, UI components, and any database schema examples already present in th

  3. 3
    Step 3

    Implement Dashboard Changes

    ▶

    Build the requested dashboard inside the existing React + Vite + Express app. Keep the Express server as the single serving process, add or update API routes as needed, and keep frontend state handling explicit enough that loading, empty, and error states render clearly. Do not m

  4. 4
    Step 4

    Build the Dashboard

    ▶

    Run the production build from the dashboard directory and capture the output.

  5. 5
    Step 5

    Smoke Test the Server

    ▶

    Start the Express server on port `3847`, capture logs, and verify the app responds. If the project has a dedicated start command, use it; otherwise use the local server entrypoint documented in the project.

  6. 6
    Step 6

    Validate the User Experience

    ▶

    Open the App tab or a browser pointed at `http://127.0.0.1:3847/`. Confirm the page is not blank, the main dashboard data loads, interactive controls work, and the layout is usable at the target viewport. Record the validation outcome in `/app/results/validation_report.json`.

  7. 7
    Step 7

    Iterate on Errors (max 3 rounds)

    ▶

    If build, server, or browser validation fails, make the smallest targeted fix and repeat Steps 4-6. Stop after max 3 rounds and write the remaining failure clearly into `summary.md` and `validation_report.json`.