Devotd Github
NewAI agents

eduwxyz/tucano v0.2.2 — A Collaboration Tool for Local Coding Agents

eduwxyz/tucano is a TypeScript application that groups multiple coding agents into channels and roles on a local computer and configures per-tool permissions and execution approvals.

Repositoryeduwxyz/tucano ↗

eduwxyz/tucano is a TypeScript application that groups multiple coding agents into channels and roles on a local computer and configures per-tool permissions and execution approvals.

GitHub description: Your team of AI agents, on your Mac, Windows or Linux. Agents that talk in channels, write code, review each other and use your tools, on the Claude Code, Codex, OpenCode or Pi you already have. Anything that touches the real world waits for your yes.

stars
★ 61
Forks
14
Language
TypeScript
License
Apache-2.0
Latest release
v0.2.2
Read original source ↗View on GitHub ↗

The latest release of eduwxyz/tucano is v0.2.2. This project is a TypeScript application that brings multiple coding agents together as one team on a local computer. It is configured so that agents converse in channels, write code, and review each other's work. It runs on Mac, Windows, and Linux, and the description says it leverages Claude Code, Codex, OpenCode, Pi, and existing model configurations. However, the provided material contains nothing newly added or changed in v0.2.2, so the specific changes in this version cannot be identified.

The core is setting a role and tool permissions for each agent and placing important executions under the user's approval. Approval is described as a structure that not only asks whether to run something but also shows the concrete task content. It is designed so that several agents take on investigation, modification, and review in turn and use tools together. Memory review and recurring task schedules, as well as a separate browser and execution environment, are also introduced as features. It is licensed under Apache-2.0 and promotes a local-first configuration that does not send conversations and files to its own server.

This evaluation is limited to the scope of features listed in the repository description and README. The provided material contains no per-release change history for v0.2.2 and no results from actual use, so there is no basis for judging how the features behave or how complete they are beyond that. Therefore, it is confirmed that this is a tool intended to handle multi-agent collaboration and per-task approval procedures in one place, but the specific extent of improvement in this release cannot be known from the material alone.