Aura IDE is a local, open source AI coding harness with Planner/Worker agents, visible write status, safe file edits, and Craft code quality checks. Shape the workflow, swap providers, and keep control of the version that works for you.
No reviews yetBe the first to leave a review for Aura IDE
Maker
📌
I built Aura because I kept losing trust in AI coding workflows.
A lot of us find a setup that clicks, build our habits around it, and then one day it changes. The model feels different. The workflow shifts. The tool you trusted yesterday feels unfamiliar tomorrow.
Aura is my attempt to make that less fragile.
It’s a local, open-source AI coding harness built around visible workflow state: what was planned, what the Worker changed, what actually applied, what failed, and why.
The bigger idea is ownership. You can use Aura to shape the AI coding workflow you actually want: tune the Planner/Worker loop, refine Craft, swap API providers, and keep the version that makes sense for you.
If a provider changes, you can switch. If the workflow needs to evolve, you can modify it. If a version feels right, you can keep it.
The latest release also adds a proper Windows installer and in-app update path, so Aura can now be installed and updated like a normal desktop app.
I’d love feedback from developers using AI coding tools:
- Does the workflow make sense?
- What feels confusing?
- What would you need before trying it on a real repo?
Report
Maker
Quick Aura update:
I used the latest Aura release to build Signal Desk, a local lead discovery and outreach tool.
Love the focus on local-first and open-source here. The Planner/Worker agent split is exactly how complex, multi-step coding tasks should be handled. But what really catches my eye is the "safe file edits" and visible write status—there is nothing worse than an LLM blindly nuking a file or hallucinating its way through a repository. Being able to swap providers and avoid vendor lock-in makes this a massive win. Definitely going to spin this up locally!
Really appreciate this, and huge congrats on your launch today.
That trust piece is exactly why I built Aura. I wanted an AI coding workflow where the plan, file edits, write status, and validation are visible instead of hidden behind a black box.
Quick Aura update:
I used the latest Aura release to build Signal Desk, a local lead discovery and outreach tool.
The workflow was:
spec → phased build → Worker runs → validation → finished MVP
Signal Desk scans sources, scores relevant conversations, writes inbox cards, and helps draft replies.
The surprising part: the full build ran on cheap models and cost about ten cents in model usage.
I wrote up the workflow here:
https://aura-ide.hashnode.dev/ho...
mailX by mailwarm
Love the focus on local-first and open-source here. The Planner/Worker agent split is exactly how complex, multi-step coding tasks should be handled. But what really catches my eye is the "safe file edits" and visible write status—there is nothing worse than an LLM blindly nuking a file or hallucinating its way through a repository. Being able to swap providers and avoid vendor lock-in makes this a massive win. Definitely going to spin this up locally!
@manal_essalek1
Really appreciate this, and huge congrats on your launch today.
That trust piece is exactly why I built Aura. I wanted an AI coding workflow where the plan, file edits, write status, and validation are visible instead of hidden behind a black box.