If you've used Bolt.new, Lovable, or v0
Continue your app with a local coding workflow.
If you've built an app with Bolt.new, Lovable, or v0 and want to work on it locally, aidd can help you plan changes, run the coding assistant you choose, and review what it did.
Moving to a local workflow
Once you have the project's source files on your machine, you need to get the app running, decide what to change, and check that those changes work. aidd gives you a place to keep that plan and manage the coding runs.
This is a choice of workflow. If your current builder suits the project, you may not need aidd. It is useful when you want to manage the source locally, choose your coding assistant, or use the same process across several apps.
What aidd is
aidd runs on your own computer and manages AI coding work. You describe what you want in plain language through a chat or guided interview. It keeps your answers with the project, turns the plan into a feature checklist, and hands the work to your chosen coding assistant. The assistant is instructed to run the project's checks and report the results.
You can start from scratch, use a template, or point aidd at an existing app, including source you've exported from an app builder. It reads what's there and plans around it. This is not an automatic migration: the app may still need dependencies, environment settings, and services configured to run locally.
Plan
Your idea becomes a spec and a checklist of features, in order.
Build
An AI assistant works through the checklist in steps you can watch.
Verify
The assistant runs the project's checks and reports what passed or failed.
Record
Run history keeps the logs, feature status, and any reported usage data.
Run history and diffs help you investigate regressions and review what the agent changed. aidd flags blocked work for review; passing checks do not guarantee the app is free of bugs.
What changes for you
Local project files
The project and its aidd records live on your machine as ordinary files. Use your own editor and keep those files if you stop using aidd.
Reviewable changes
Work is tracked against a feature plan. The run logs and code diffs give you a starting point for reviewing each change and checking for unintended effects.
Context for later work
The spec, feature list, and history stay with the project and provide context for later runs. You can update the plan as the app changes.
Built-in inspections
Built-in audits ask the agent to inspect security, performance, and code structure, then record evidence-backed findings as tasks to review.
Still conversational
A built-in Director chat suggests work and can start it for you. The Telegram bridge lets you use it from your phone after you connect your own bot and allow-list your chat.
Grows with the project
One app or a dozen, the same dashboard shows every project, what's in progress, and what needs your decision.
What you'll need
aidd ships as source rather than a double-click installer. You install Git and Bun and run it from a checkout. Windows 11 is the first-class, smoke-tested platform; Linux runs the whole test suite automatically, and macOS is expected to work but is only checked by hand. There's no hosted preview URL - the panel can start and stop your app locally, but where it runs and how it deploys is your call. And you connect an AI you already have access to: an agent CLI like Claude Code or Codex, an OpenAI-compatible API, or a model running locally through Ollama or LM Studio.
Local records, your choice of backend. aidd stores project
metadata, run history, settings, and usage data locally. It has no account or
hosted control plane and sends no telemetry. The panel listens on
127.0.0.1 by default. Coding runs send prompts and relevant project
content to your chosen backend, which may be a cloud service or a local model.
Configured integrations and downloads you initiate can also make network
requests.