Community Contribution Workflow
Contribute to Agent Hub
Create or improve Agent Hub packs for the tools you care about using your local AI coding agent, then contribute the result through GitHub in the open.
Agent Hub packs are Markdown guides that help AI coding agents use libraries, APIs, SDKs, and product surfaces more accurately.
How It Works
- Pick a pack you want to improve or a project you want Agent Hub to cover.
- Use your local AI coding agent plus the repo prompts to draft, revise, and validate the pack.
- Open a GitHub PR so the change can be reviewed, managed, and merged in the open.
Ways to Contribute
Create or Improve Packs
Start a new pack for a library, API, or product you know well, or tighten an existing pack by updating versions, adding workflows, fixing structure, or improving troubleshooting guidance.
Local AI Workflow
Use Your Local AI Coding Agent
Use the prompts and runbooks in this repo to draft a new pack or improve an existing one with your local AI coding agent. Then validate the result locally and open a pull request.
The pack tutorial shows which repo documents to trust and how to move from draft to a review-ready contribution.
Community-Generated Packs
All Content Stays in the GitHub PR Flow
Agent Hub knowledge packs are meant to be community generated, versioned, reviewed, and improved over time. The repo is the source of truth for pack content, prompts, tutorials, and validation rules, so contributions land through pull requests instead of disappearing into one-off prompt experiments.
What You Can Add
- New packs for libraries, frameworks, APIs, SDKs, and product surfaces
- Better workflows, troubleshooting sections, and version guidance
- Prompt and evaluation improvements that raise the quality bar for future packs
How Review Works
Agent Hub AI agents help monitor pack quality through the same GitHub PR workflow contributors already use. They can flag gaps, validate structure, and support iteration, while humans still review and merge the final change.