// docs
Get it running.
Short walkthroughs for the things you do once — create a project, connect a server, hand work to the agent.
GET STARTED
Create a projectGo from nothing to a running application on your own server, with HTTPS and a real domain, in about five minutes.Connect a serverEnroll a box you own with a one-time install command — no SSH needed after that, and Weaveflow never holds your keys.Connect your integrationsWhat each connection unlocks, which ones the create flow needs, and how your credentials are stored.
WORKING WITH THE AGENT
File your first ticketWrite a ticket, assign it to the agent, and watch it work — from your phone or anywhere else.Review and approve the agent's workThe report, the diff, the questions — and how approval turns into a pull request in a repo you own.Use the live previewEvery project can run its work-in-progress on a real URL, so you can watch the agent's change in a browser before approving it.Work from your own editorThe dashboard is optional at your desk — clone the repo, use your own tools, push, and everything stays in sync through git.
RUNNING IN PRODUCTION
Environments and secretsPer-environment config, write-only secrets, and changes that apply to the running app in seconds without a rebuild.Deploys, rollback and failuresPush-to-deploy, one-click rollback, and what actually happens when a deploy breaks.Attach your own domainPoint a domain you own at your project — the record to create, how to verify it, and why the free address never goes away.Weaveflow API and agent accessThe Weaveflow API base URL and authentication, and how AI agents should read this site.