dardagreg

I think the upside for everyday people outweighs the (current) risks. I've been using Harper (harper.new) to keep track of my (complex) medical info. Obviously one of the use cases of AI is pulling out data from pdfs/images/etc. This app does that really well so I don't have to link with any patient portals. I do use the AI chat sometimes but mostly to ask questions about test results and stuff like that. Its way easier than trying get in to see my doc

I've been using a tool called Fabrk AI to build agents that have tools that integrate with any service that has Oauth. I've hooked it up to github, slack and gmail for now.

I write issues in a repo and the agent pulls it down, creates a branch and writes the code. Works pretty well. Even integrates with e2b.dev and can run code.

This is the template I started with: https://www.fabrk.ai/agent/github-demo

That makes sense. Like it. One question - in the video you mentioned adding credentials that would be available to the function calling. I didn't see that tool when I forked the HN demo agent. Where can I find that? I really like the credentials idea. Having a way to use them for tool/function calling but not available to the AI is smart. With credentials you could envision giving the AI access to all sorts of services. Really opens up the possibilities. Let me know where I can find the credentials tool. Thanks

How does this work? I'm assuming you're using openai's function/tool calling. But how do you support the forking of the tools? The scheduling is cool.