Skip to content
Neural Daily

ToolsNews

SuperPlane 1.0: open source software factory for agents

🇭🇺 Olvasd el magyarul →

The Semaphore team released SuperPlane 1.0: an open source, model-agnostic system that hands selected backlog tasks to coding agents, returns the work with evidence, and leaves final approval to a human.

By: Mikola Máté Patrik · · 2 min read

SuperPlane 1.0 announcement graphic: the open source software factory for teams
Image: SuperPlane · Press kit · Official social image from the SuperPlane 1.0 launch post · Source

What was released?

The Semaphore CI team released SuperPlane 1.0 (SuperPlane). It is not another coding agent but the process around one: it decides what the agent should get, runs it, then checks what it returned. It is Apache 2.0 licensed, self-hostable, and works with any model or agent.

How it works

  • It assesses the backlog, scoring each task on clarity, complexity and verifiability to show what can be delegated to an agent.
  • It runs the coding agent in a cloud development environment.
  • It returns evidence: executed checks, UI screenshots, recordings.
  • When CI fails or a review comment arrives, it loops the task back to the agent and hands only finished work to a human for approval.
  • A shared kanban board shows what is running, and a Velocity page measures end-to-end throughput.

Numbers

The team says 52% of its own merged PRs last month went through this process. That is internal data from one team, not an independent measurement.

What to keep in mind

The company suggests starting with small bug fixes, dependency updates and security updates. These are well defined, easy to verify and low risk. Approval always stays a human step.

Why it matters

Coding agents today mostly run on one developer's terminal, with no team-level view. SuperPlane turns backlog triage, CI feedback and evidence collection into a shared open source layer you do not have to build in-house.

    Produced with the help of AI, reviewed by an editor. How we make our articles

    Reviewed and edited by the Neural Daily desk. About

    Related articles