Orquestrador Maestro CLI
A process kit for agents: a skill router, project memory in a DEV folder and result verification before saying done.
- 63 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- Changes client settings, connects hooks and writes state files into the project.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/orquestrador-maestroDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Changes client settings, connects hooks and writes state files into the project.
This is third-party code. Review the repository files before installing.
What it does
Orquestrador Maestro sets up the environment so an AI agent follows the same process across tools: it reads global and project instructions, opens only the needed context, works within limits, verifies the result and leaves a trail for the next session. Project state lives in a `DEV/` folder: goal, decisions, verification and next step. A router picks one main skill out of hundreds by intent, complexity, stack and touched files. Integrations exist for Codex, Claude Code, OpenCode, Cursor, Gemini CLI, Windsurf, Kimi Code and other clients. Codex Desktop and ChatGPT get a plugin with hooks and a local MCP, and the `install`, `verify` and `update` commands manage setup.
Who it is for. Developers who work on one project across several AI tools and are tired of re-explaining context every session.
Good fit when
- A project moves between Codex and Claude Code and needs shared rules and memory.
- The agent opens too many irrelevant files and burns context.
- A task may close only after verification, not on the model's claim.
- A new session starts from the `DEV/` folder with the goal and next step recorded.
- A team measures context savings on its own machine with the built-in benchmark.
Not a fit when
- A single quick edit needs no process and no state record.
- Changing client configuration and hooks is not allowed.
- Docs are needed in English: the main repository text is in Portuguese.
Example request
Use Orquestrador Maestro: read the project instructions, open only the needed context, solve the task, verify the result and record the state in `DEV/`.Limitations
Node.js 20.19 or newer is required. The main documentation is in Portuguese and the English README is shorter. The kit changes client settings and connects hooks: Codex Desktop asks for trust confirmation, and it is wise to try the installation in dry-run mode first, with rollback described in the guide. Each tool stays responsible for its own login, model and keys. The stable line is V1 1.1.1, and older versions differ in router and behavior.
How to disable. Run `orquestrador-maestro desktop-plugin status` and disable the plugin in client settings. Delete the `DEV/` folder in projects or keep it as documentation, and remove global instructions following the installation guide.
FAQ
How to choose a skill?
You need not memorize names: the router picks one main skill and adds supporting ones only when justified. A selection guide sits in `docs/skills/choose.md`.
How to install the Codex Desktop plugin?
With `orquestrador-maestro desktop-plugin install`, then `status` and `doctor`. After restarting the client enable the plugin and approve the hooks.
What does the DEV folder hold?
The project goal, decisions, verification and next step. People and other tools can read the files, so switching agents does not break the work.
How to check context savings?
The repository has a benchmark harness with agent and verifier Docker images. The procedure is in `docs/benchmark.md`.
How to update?
Run `orquestrador-maestro update`, then `orquestrador-maestro verify`. The stable V1 ships in the npm `latest` channel.
Discussion
No questions or reviews yet.
Related
CLI toolsOne agent launches others in the background: Claude Code, Codex, Copilot CLI, Grok Build and Pi, with a model chosen per stage.
A desktop app where Claude Code, Codex, Gemini and other agents work as a team of specialists in a shared project
A skill with 25 commands to audit and improve the prompts, tools and agent designs of a project in Cursor, Claude Code and Codex
Skills and the porch CLI for agent orchestration: a strong model plans, cheaper ones code, several model families review.
A bridge between local CLI agents: Codex, Cursor or Claude Code delegate work to Antigravity, Grok, Kimi and seven more.
An orchestration plugin for Claude Code, Codex, Gemini CLI and Qwen Code: 39 specialists, parallel subagents, persistent sessions and standalone audits
README in short
The project runs from installation to benchmark: hundreds of skills, a router, `DEV/` memory, a desktop client plugin, workflows and Portuguese and English docs.