Devotd Github
NewAI agents

orafaelfragoso/crewyard v0.6.0 — Local Orchestration Tool for AI Coding Agents

Introduces the latest GitHub release v0.6.0 of orafaelfragoso/crewyard and its capability for orchestrating task assignment, review, and merging across independent coding agents.

Repositoryorafaelfragoso/crewyard ↗

Introduces the latest GitHub release v0.6.0 of orafaelfragoso/crewyard and its capability for orchestrating task assignment, review, and merging across independent coding agents.

GitHub description: Local-first coordination engine for independent AI coding agents: task leases, isolated worktrees, versioned context, reviews and a safe merge queue for Codex, Claude Code and OpenCode.

stars
★ 2
Forks
0
Language
TypeScript
License
MIT
Latest release
v0.6.0
Read original source ↗View on GitHub ↗

orafaelfragoso/crewyard is a local-first orchestration tool that connects the planning, task assignment, execution, review, and merging of multiple independent AI coding agents. The latest GitHub release is v0.6.0. However, the collected materials contain no changelog for this version, so no newly added or modified features can be confirmed. A task lease refers to the procedure through which an agent secures the work it has taken on for a set period of time. The tool is designed to divide agents' work using task leases and isolated Git worktrees.

Crewyard goes beyond assigning tasks and handles context management, review, and integration in a single flow. It provides version-pinned contexts, links supporting evidence to reviews, and is designed so that merges are handled sequentially from a queue. It offers a command-line tool and an MCP server, and stores its state in a local SQLite database. Accordingly, the fact that it does not require a hosted Crewyard server is the core of its independent-agent orchestration model. It is written in TypeScript and licensed under MIT.

The README states that the project is still in a pre-1.0 development stage and that commands and settings may change. The specific changes in v0.6.0 are likewise absent from the materials, making it difficult to judge the scope of this release's improvements. What can be confirmed is the design of the orchestration flow and its components. So rather than evaluating release-by-release changes or actual results of use with the current materials alone, it is closer to a tool worth examining together with the local orchestration approach and the mutability of its development stage.