My AI wrote API docs. Then a parser fact-checked it.

A solo developer built Docloom, a tool that uses an LLM to write API documentation while a separate parser fact-checks the generated content against the actual codebase, preventing the model from inventing endpoints. The tool presents a diff of changes after each merge and requires human approval before anything goes live, following the developer's experience with an LLM that confidently documented a nonexistent endpoint. Docloom is free to try and the developer is soliciting sample repositories to test it against.

Let me tell you how API docs die. Nobody decides to let them rot. You ship a feature on Tuesday, promise yourself you'll update the docs on Wednesday, and by Friday something else is on fire. Six months later your docs describe an API that doesn't exist anymore. The obvious fix is to let an LLM write them. So I tried that. It read my code and gave me something confident, polished, and partly made up. It documented an endpoint that wasn't there. That was the moment the problem changed shape for me. The AI wasn't bad at writing. I had just given it two jobs: figure out what my API does, and explain it. The explaining, it's great at. The figuring out should never be a guess. So I split the jobs That's the whole idea behind Docloom. The AI describes. It never discovers. Think of it as a writer who's only allowed to write about facts a fact-checker already signed off on. You stay in control When you regenerate after a merge, Docloom shows you a diff of what changed. Nothing goes live until you look at it and approve it. Docs you never reviewed are how this problem started, so I didn't want to rebuild it. Yes, I know you're handing over a repo That's a big ask, so here's exactly what it gets: It's free, no credit card. If you want to try it: https://docloom.babar-wealthpilot.workers.dev https://docloom.babar-wealthpilot.workers.dev I'm building this solo, so I mostly want to find out where it falls over. Tell me your stack in the comments and I'll run it on a sample repo and share what happens, good or bad.