what is devagent-studio fr?

DevAgent Studio is an open-source workbench built to help software teams understand, review, and keep track of their projects, rather than to write code for them. It brings together several specialized AI agents, one that plans work, one that analyzes a project's structure, one that reviews code for risks, one that manages project knowledge, a supervisor that coordinates the others, and one that writes reports, all working together on a task and leaving a traceable record of what happened. The tool is explicitly not a coding assistant that writes your software. Instead it focuses on making the process of building software easier to inspect, audit, and improve: scanning a codebase to summarize its structure and flag risks, reviewing code changes with a mix of fixed rules and AI judgment, and maintaining a searchable knowledge base about the project that can be filtered by who is allowed to see what. It also keeps a governed long-term memory of useful facts and preferences learned from conversations, but only adds anything to that memory after a person explicitly confirms it, and it tracks when a new fact replaces an older, conflicting one. Workflows in DevAgent Studio are built visually, by dragging and connecting steps on a canvas, and that visual diagram becomes something that actually runs rather than staying a picture. Sensitive actions can be gated behind a human approval step, and a task can be checkpointed, resumed, or retried if something goes wrong partway through. The tool also tracks how AI models are being used across the system: which model each agent calls, the cost and speed of each call, fallback behavior when something fails, and side-by-side comparisons between different configurations. It is built with Python on the backend, using FastAPI and a workflow framework called LangGraph, with a React interface on the front end, and it can run third-party code in an isolated Docker container with no network access as a safety measure. The interface is available in both English and Chinese and offers two visual themes to choose from. This project is aimed at engineering teams and technical leads who want more visibility, auditability, and control over how AI tools are used inside their software development process, rather than developers looking for a tool that writes code directly.