DEDUPED Hot AI repos. Real use cases. What people build. Get the Monday 5

How to spin up a working recommender in one command with Gorse

Run the playground image, which imports GitHub repository data and trains recommendations you can explore.

Short answer: Start the playground with docker run -p 8088:8088 zhenghaoz/gorse-in-one --playground; then open the dashboard at http://localhost:8088; wait for the Generate item-to-item recommendation task to complete on the Tasks page. All steps.

  • Top pick
  • Data and analytics
  • Beginner
  • 20 minutes
  • From the README
  • Claude Code
  • Codex
  • Cursor
  • Gemini CLI

Who it is forA developer evaluating whether a self-hosted recommender fits their product.

The problemSetting up a recommender usually means data pipelines and model training before any demo.

The outcomeA running Gorse instance with a dashboard at localhost:8088.

What you need first

  • Docker for the gorse-in-one image
  • curl or any HTTP client for the REST API
  • For production: MySQL (MariaDB), MongoDB or Postgres for storage, with a cache such as Redis

Steps

  1. Start the playground with docker run -p 8088:8088 zhenghaoz/gorse-in-one --playground
  2. Open the dashboard at http://localhost:8088
  3. Wait for the Generate item-to-item recommendation task to complete on the Tasks page
  4. Browse items, users and recommendations in the dashboard

Prompt for your agent

Using gorse-io/gorse, run `docker run -p 8088:8088 zhenghaoz/gorse-in-one --playground`, wait for the item-to-item task to finish, and describe what the dashboard shows.

Caveat. Playground mode downloads data from GitRec on startup, so it needs internet access.

Use this in your AI agent

Ask your agent

Using the Deduped MCP, get the setup steps for gorse-io/gorse and "Spin up a working recommender in one command"

Add Deduped to Claude Code once

claude mcp add --transport http deduped https://dedupedai.com/mcp

Cursor, ChatGPT, others

Share: X LinkedIn Hacker News Email Copy link

Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer