What it does

LINKIT is a web-based, drag n' drop Chainlink job spec editor with an API that enables you to simulate job spec execution in the browser.

Note: LINKIT is an existing project that was submitted to the previous Chainlink hackathon. The feature additions that have taken place as part of this hackathon are as follows:

  • Integration with OpenAI to enable the generation of sections of your job spec pipeline with natural language
  • The ability to import an existing job spec (great way to help visualize specs taken from the official docs)
  • User login and saving/loading of job specs
  • UI upgrades and dark/light theme toggle

Challenges we ran into

While I'm proud at the level of improvement I've been able to make, other commitments have meant I wasn't able to add as much functionality as I would have liked to. As as result, the user profile features are very far from where I would like them to be, and I haven't had the time to upgrade to v2 of the Chainlink node software, or support all of the available job-type-specific variables.

Accomplishments that we're proud of

The AI generation of tasks is surprisingly good given that I haven't yet trained a custom model but am instead providing as much information as can condense in the limits of a request to Open AI's APIs.

What we learned

This was my first time using the Open AI APIs, Cockroach DB, thirdweb's Auth module, shadcn's UI component library builder, so I learned a lot about those.

What's next for LINKIT

With more time and resources I would love to fully flesh out the features which are half-complete and bring a solid app to production so people can make use of it!

Built With

  • cockroachdb
  • go
  • next.js
  • openai
  • react
  • thirdweb
  • vercel
Share this project:

Updates