I built an async, self-improving AI coding agent that never leaves my machine. Now it runs in my browser, on my phone, and learns from its own work.
Every AI coding tool I tried had the same fine print. It runs commands on your machine. It reads files on your machine. It feels local. Then you check where the model lives, and it's on somebody else's server. Every file it reads, every command it runs, every line of your code takes a trip there first. So I built Flash . It's an AI coding agent where the tools run on your machine and so does the…
Flash is an innovative AI coding agent designed to operate entirely on a user's machine, without sending any data to external servers or requiring an API key. Created by the author, Flash utilizes Ollama as its model and runs on the user's hardware, allowing users to access its coding capabilities directly from their browser or phone.
The tool is particularly notable for its performance on local models, which are often slower than cloud-based alternatives. Flash addresses this issue by changing files by specifying only the necessary text changes, rather than rewriting entire files. This approach reduces the amount of work the model needs to perform, enabling the agent to operate efficiently even on slower hardware.
Flash also features an intuitive user interface, with keyboard shortcuts and a "permission prompt" that replaces traditional file editing prompts, ensuring users have full control over their actions. Furthermore, Flash offers a range of extensions that can be installed via GitHub repositories, allowing users to customize the tool's functionality to suit their specific needs.
Finally, Flash emphasizes privacy and security, with all requests originating from the Flash application itself and state changes limited to JSON data, preventing unauthorized access or manipulation.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.