DocsAlot turns scattered help center articles, knowledge base, and developer docs into one source of truth for humans and AI agents. It includes hosted MCP, llms.txt, and skill.md. Your docs show up in AI answers, onboarding gets faster, and agents stop reading stale context.
This is the 4th launch from DocsAlot. View more
DocsAlot MCP Connector
Launched this week
Create and update your knowledge base from Claude, Codex, or any compatible MCP client. Ask DocsAlot to add a page, fix an error, or improve your help center. It finds the relevant content, makes changes, creates a new version, and publishes, all from your existing workflow. See what questions users ask in your help center, discover what’s missing, and use those insights to improve your docs and decide what to build next.
Payment Required
Launch Team / Built With
Framer AI AgentsDesign and publish professional sites with AI
Promoted
Congrats on the launch, Faizan and team! I had a great time getting to know you for our recent interview on the APIs You Won't Hate Podcast. You've got a great perspective on the problem space - I was surprised how much I learned from you during our short chat.
@irreverentmike1 Thanks so much Mike. Really appreciate your support, and you were a great host, I had a lot of fun. You extracted a bunch of internal secrets out of me as well 😅.
Looking forward for the pod to be out.
@new_user_2790a57d4d It's out! Check the link above :)
@irreverentmike1 oh ofcourse 😊
the "creates a new version, and publishes, all from your existing workflow" part is the one thing I'd want a safety rail on. we keep our own internal docs in a similar loop where Claude proposes edits, and the failure mode isn't that it writes something wrong, it's that it writes something confidently wrong and nobody reviews it before it goes live because the whole point was to skip a step. does DocsAlot publish straight from the MCP call, or is there a draft/approve step before a Claude-made edit actually replaces what's live for your end users?
@galdayan Thanks Gal. So we have a two step process, as shown in the 2nd video. First one is PUSH, which saves a version on the CMS. This allows the human to share that version with other people in the team. Once everything is ready, only then someone can PUBLISH.
And every version is retained historically. And you can easily revert back by changing the published version.
@new_user_2790a57d4d that answers it cleanly, push/publish as a real gate rather than a single auto-commit step. the version history + revert is the part that actually matters to me though, since that's what turns "the AI published something wrong" from a slow-motion incident into a one-click rollback. are reverts scoped per-page or does rolling back one page risk touching others that referenced it
Fixing a doc by just asking in the tool you already work in is the part I like. Most people skip small doc fixes because opening the editor feels like a chore.
@christian_becerra2 thanks appreciate it.