Skip to main content
Docs/Agent Mode
DocsAgent Mode
Free

Agent Mode

Describe what you want. For a big request, GenMB plans the work, builds it step by step, and fixes errors on its own. There is nothing to switch on.

What is Agent Mode

Agent Mode is how GenMB builds bigger apps step by step on its own. Instead of generating code in one pass, the AI breaks your request into a series of tasks and works through them in order: planning, building, checking, and fixing errors automatically.

The AI analyzes your prompt and creates a structured task plan before writing any code. Tasks are executed one by one, with each step building on the results of the previous one. Errors are checked and fixed between steps within the same run, so no manual intervention is required. The final result is a complete, working app that matches your original prompt.

Agent Mode runs on every plan and costs 8 to 40 credits a build, based on size. There is no button or setting: GenMB decides by itself when a first request is big enough to need it.

When GenMB Uses It

GenMB uses Agent Mode for big first requests that would be hard to build in one pass: several features, screens, or parts that depend on each other. Just describe your app and GenMB picks the right approach.

  • ●Complex multi-feature apps: e.g., "Build a project management tool with kanban boards, team members, and a dashboard."
  • ●Apps with many screens: e.g., "Create an e-commerce site with product listing, cart, checkout, and order history."
  • ●Apps that need multiple iterations: Agent Mode refines as it builds, catching and fixing issues along the way.

Small requests stay fast

For simple, single-page apps, GenMB builds in one quick pass instead. You do not need to choose: describe what you want and GenMB decides.

How It Works

Agent Mode follows a structured pipeline from prompt analysis to final delivery. The AI handles every stage autonomously so you can sit back and watch your app come together.

1

AI analyzes your prompt

The AI reads your description, understands the requirements, identifies features, and determines the overall app structure before writing a single line of code.
2

Creates a task plan

Your request is broken into discrete, ordered tasks, for example, "Set up navigation", "Build dashboard page", "Add data layer". You can see the full plan before execution begins.
3

Executes tasks sequentially

Each task generates or modifies code, building on what was created in previous tasks. The AI maintains context across the entire session.
4

Checks and fixes errors

After each task, the AI checks for errors and fixes them before moving to the next task. This check-and-fix loop means fewer bugs in the final result.
5

Delivers the final app

Once all tasks are complete, the finished app is ready for preview and deployment. Every file is saved and a version snapshot is created automatically.

Watching Progress

Agent Mode provides real-time visibility into what the AI is doing. You can monitor each step and intervene if needed.

  • ●The task plan is displayed in the UI before execution begins, you can see what the AI intends to build.
  • ●Each task shows its status in real time: pending, in progress, completed, or failed.
  • ●You can watch code being generated and modified as each task executes.
  • ●Cancel at any time, the app preserves whatever has been built up to the cancellation point.
If the AI is heading in the wrong direction, cancel early and tell it in chat what you want instead. You keep all the work completed before cancellation.

Session Limits

Agent Mode sessions have limits to prevent runaway execution and ensure system stability. These limits are generous enough for most apps but worth knowing about for very large projects.

  • ●20 tasks per session: if your request needs more, ask for the rest in chat afterwards.
  • ●12-minute timeout per task: individual tasks that take longer are stopped.
  • ●1-hour inactivity timeout: the session ends if no task runs for 60 minutes.
You can keep building at any time by asking in chat. Every change picks up from the current state of your app.

Agent Mode and Chat Edits

After your first version, every change you ask for in chat runs the same build-and-check loop automatically. You just type what you want.

  • ●First build: a big request is built in Agent Mode, step by step, for 8 to 40 credits based on size.
  • ●Chat edits: each change is built and checked the same way, for 1 to 40 credits based on how much work it takes.
  • ●Questions are free: if you ask a question, or for a plan, an explanation or an audit, the AI answers without changing your app and charges nothing.

Just ask

Small tweaks ("move the sidebar to the right", "add a search bar") and big features both go in the same chat box. GenMB works out how much to do.

FAQs

What plan do I need for Agent Mode?▾
Agent Mode runs on every plan. There is nothing to switch on: GenMB uses it by itself when your first request is big enough and you have the credits. A build costs 8 to 40 credits, based on size.
Can I cancel an Agent Mode session in progress?▾
Yes. You can cancel at any time. The app keeps whatever was built up to that point, and you can carry on by asking for the rest in chat.
What happens if a task fails during Agent Mode?▾
The AI checks for errors after each task and fixes them within the same run. If a task fails and cannot be recovered, the session will note the failure and continue with remaining tasks when possible. The app preserves all successfully completed work.
How is Agent Mode different from GenMB Code?▾
Agent Mode is how GenMB builds a big request step by step on its own. GenMB Code is a coding assistant where you guide the AI through specific file edits.

Ready to build?

Create your first app for free, no credit card required.

AI AgentsMCP Integration