How your agent and Feldova work together.
Connect your agent
Once, in about a minute.
Add Feldova to the AI agent you already use: Claude Code, Claude, ChatGPT, Cursor or any agent that supports MCP. You sign in to Feldova and approve what the agent may do. There is no API key to copy, and you can disconnect it any time in Settings.
Ask for what your team needs
Plain words, any language.
Tell your agent about the tool you want, the way you'd explain it to a colleague: "A CRM for our agency clients." Your agent writes the app. Feldova checks every version before it saves it, and the preview in your studio updates as the agent works. Not quite right? Tell your agent what to change.
Publish and share
A real app, at its own URL.
When you're happy, ask your agent to publish. You choose who sees it: people with the link, or everyone in the Feldova directory. Your team signs in with one Feldova login, and roles decide who sees what. Hosting and the database run for you, in the EU.
Work with your data
Your agent can run the app, too.
Your agent can also read and change the data in your apps: import a spreadsheet, add this week's tasks, or fix a record. Row changes it makes show up in the studio's Data tab, where you can undo them.
Whatever your team actually runs on.
These are starting points, not a fixed menu. If your team works with it, you can build an app for it.
A pipeline for agency clients, with deals, contacts, and follow-ups.
What's in stock, what's running low, and what's on order.
Every new hire's first-week checklist, the same way each time.
Requests that route to the right person and track their status.
The numbers your team checks every morning, in one view.
The tool you've never found off the shelf. Ask your agent for it.
Feldova handles the plumbing. You and your agent shape the work.
You stay focused on how your team operates. Everything underneath is taken care of.
- Authentication and roles
- Hosting, in the EU
- The database behind every app
- Checks and versions for every change
- Email, schedules, files and AI inside your apps
- The workflow each app follows
- The fields and what they capture
- The logic and the rules
- Who sees what
- How each app fits your team
One source of truth, hosted in the EU.
Your apps and their data run on EU infrastructure.
Every app reads from the same data, so nothing drifts out of sync.
It belongs to your team and stays under your control.
Yeah, but how does it actually work?
- Which AI agents work with Feldova?
- Any agent that supports MCP connectors with sign-in. The connect page has step-by-step setup for Claude Code, Claude (desktop and web), ChatGPT and Cursor.
- What can my agent do?
- Create apps, read and change their code, check and save versions, read errors from the preview, and publish when you ask. It can also read and change the data in your apps. It only works in workspaces you belong to, with the access you approved.
- What can't my agent do?
- It can't change billing, invite people or see workspaces you don't belong to. Row changes it makes show up in the studio's Data tab, where you can undo them, and you can disconnect it any time in Settings → Connected agents.
- What runs on Feldova?
- Your apps: hosting, login and roles, the database, file uploads, email, schedules, and AI features inside the apps. Your agent does the building, and Feldova runs the result. Hosting and the database are in the EU.
- Do I need to be technical?
- No. You ask in plain words, in any language. Your agent writes the code, and Feldova checks it before it goes live.
- Where does our data live?
- In the EU, in one place every app reads from. It's yours.