Prioritizing unifying docs and building doc integrations with Popular coding assistants
As more and more coding assistants get even more popular, I think tools like Convex should prioritize developing easier ways for documentation to integrate with coding assistants.
I only suggest this, because I genuinely want Convex to win! Having extensively used all other offerings, from early adoption of prisma, to supabase, drizzle, trpc, grpc, REST, years of graphql(still presently for day job), I think Convex has a real shot at winning and I am completely rooting for you.
The work you've done with the application, customer support, docs, guides, youtube, stack, hackathons, HIPAA and SOC2, have all been awesome. Love it all.
However, we move with the future. And the future is - I need to hire a new Engineer. He has not worked with Convex - proof from my several failed attempts at hiring Engineers with some knowledge both from the jobs channel and upwork. Said Engineer needs to ramp up quickly on Convex, and has to spend several days going through the docs, finally finding convex-helpers, breaths a sigh of slight relief. After several PR feedback, finally starts reading the Stack docs I share, etc.
Said Engineer already has Cursor, or Replit, or ClaudeDev, etc. Decides to ask, and it gives it the most inefficient query ever. Probably wrong. Wrangles with it. Puts up a PR, and we spend the next few days going back and forth.
Better future: Said Engineer setups up the project, uses working current docs integration that bring in the juice from the awesome guides on Stack with convex helpers, chat does a much better job. PR feedback is shorter. Everyone is happy.
I did setup the docs with Cursor, and it is still not great. Spent a few hours trying to scrape Convex docs for Replit to setup for a new engineer, with a short guide from the Replit team, no luck.
At the end of the day, the ability to easily hire people for your stack is a key factor in adoption.
Some things that would be great to see:
- Continued in comment:
I only suggest this, because I genuinely want Convex to win! Having extensively used all other offerings, from early adoption of prisma, to supabase, drizzle, trpc, grpc, REST, years of graphql(still presently for day job), I think Convex has a real shot at winning and I am completely rooting for you.
The work you've done with the application, customer support, docs, guides, youtube, stack, hackathons, HIPAA and SOC2, have all been awesome. Love it all.
However, we move with the future. And the future is - I need to hire a new Engineer. He has not worked with Convex - proof from my several failed attempts at hiring Engineers with some knowledge both from the jobs channel and upwork. Said Engineer needs to ramp up quickly on Convex, and has to spend several days going through the docs, finally finding convex-helpers, breaths a sigh of slight relief. After several PR feedback, finally starts reading the Stack docs I share, etc.
Said Engineer already has Cursor, or Replit, or ClaudeDev, etc. Decides to ask, and it gives it the most inefficient query ever. Probably wrong. Wrangles with it. Puts up a PR, and we spend the next few days going back and forth.
Better future: Said Engineer setups up the project, uses working current docs integration that bring in the juice from the awesome guides on Stack with convex helpers, chat does a much better job. PR feedback is shorter. Everyone is happy.
I did setup the docs with Cursor, and it is still not great. Spent a few hours trying to scrape Convex docs for Replit to setup for a new engineer, with a short guide from the Replit team, no luck.
At the end of the day, the ability to easily hire people for your stack is a key factor in adoption.
Some things that would be great to see:
- Continued in comment:
