convez.ai
AgenteAutomaçãoIntermediário
Orkas-AI

Orkas — Desktop Multi-Agent AI Commander

Orkas-AI/Orkas

Orquestre equipes de agentes IA local-first, sem código, com seus próprios LLMs

Ver no GitHub
BomPluginScore · de 0 a 100

Aplicativo desktop open-source para construir e orquestrar equipes de agentes IA especializados através de uma única conversa. Funciona local-first (seus dados, chaves de API e conversas nunca saem da sua máquina), suporta múltiplos provedores de LLM (Claude, OpenAI, Gemini, DeepSeek, etc.) e permite que agentes evoluam autonomamente. Disponível para macOS, Windows e Linux.

Instalação

$npx convez-ai Orkas-AI/Orkas

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
9,4 mil
Forks
561
Acompanhando
220
Saúde
42%
Último commit
há 23 horas
Versão
v2026.11.16
Issues abertas
19
Contribuidores
2

Casos de uso

  • Automatizar relatórios e pesquisas de mercado com sub-agentes especializados
  • Quebrar especificações de produto em tarefas distribuídas entre agentes
  • Analisar documentos locais mantendo dados privados na máquina
  • Orquestrar agents CLI existentes (OpenClaw, Claude Code, etc.) em workflows

Veredito

Orkas preenche um espaço único entre frameworks de código (CrewAI, LangChain) e plataformas SaaS: oferece orquestração multi-agente sem código, mantendo total soberania de dados. Ideal para profissionais que querem automação complexa com privacidade e flexibilidade de provedores de LLM.

README

A tradução para o português deste README ainda está na fila.

Orkas

Open-source, local-first desktop app to build and command a team of AI agents — in one chat, with your own LLM keys. Fully offline-capable. macOS · Windows · Linux.

License: MIT Stars Platform Download X: @leochenpm

English · 简体中文

Orkas demo

A commander LLM assembles a team of sub-agents and dispatches them in parallel or in series. Agents self-evolve through reflection. Your conversations, files, and API keys never leave your machine.


What is Orkas?

  • What it is — a desktop GUI app where you build a team of specialized AI agents and command them through a single chat. Not a code framework, not a hosted SaaS.
  • Local-first — conversations, files, API keys, knowledge bases, and custom agents all stay on your disk. Model calls go straight from your machine to the provider — never through Orkas servers.
  • Multi-agent (lead + sub-agents) — one commander LLM dispatches sub-agents in parallel or in series; each agent has its own private skills and memory, and self-evolves after each task.
  • Bring your own LLM keys — plug in Claude, OpenAI, Gemini, DeepSeek, Kimi, GLM, Qwen, MiniMax, or Doubao. Mix providers across agents. No vendor lock-in.

⭐ If Orkas is useful to you, a star helps more people find the project.


What can you build with it?

  • Automate recurring reports & market research — a sub-agent that gathers, summarizes, and ships a weekly report.
  • Turn a product spec into dev tasks — the commander breaks a PRD into tasks and dispatches them across agents.
  • Chat with your documents & run local data analysis — drop files in, keep the data on your machine.
  • Orchestrate existing CLI agents — OpenClaw, Hermes-Agent, Claude Code, Codex and other local CLI agents plug in as backends and take handoffs.

Download

  • Get the app → orkas.ai (macOS · Windows installers)
  • Run from source → see Quick start below

How Orkas compares

ToolWhat it isHow Orkas differs
LangChainA developer framework/library for building LLM apps and agents — code-first, embedded in your own Python/JS app.Orkas is a no-code desktop GUI: you assemble and command a team of agents through chat, not by writing orchestration code. Data and keys stay local by default.
CrewAIA Python framework for orchestrating role-playing autonomous agents — you define crews and agents in code.Orkas gives you the same multi-agent orchestration without code, in a desktop app, with local-first storage and per-agent self-evolution built in.
Cloud agent platforms (SaaS orchestrators)Server-hosted; conversations, files, and API keys live on the vendor's infrastructure.Orkas is local-first: everything stays on your machine, and model API calls go straight to the provider — never archived by Orkas.
OpenClawA single always-on personal assistant reaching you across messaging channels.Orkas builds a team of specialized agents directed from one desktop chat — and OpenClaw plugs in as an Orkas CLI backend.
Hermes-AgentNous Research's self-improving personal agent (TUI + multi-channel gateway).Orkas is desktop-GUI and team-shaped, with per-agent private skills and meta-cognition — and Hermes-Agent plugs in as an Orkas CLI backend.

Orkas is for you if you want a team of agents (not one assistant), a desktop GUI with file drop-in and visual agent management, and your data, keys, and agents on your own disk rather than a vendor cloud.


Quick start

Prefer the packaged app? Use the production installer links:

To run from source:

Requirements: Node 20+ · Python 3 · macOS / Windows 10+ / recent Linux

git clone https://github.com/Orkas-AI/Orkas.git
cd Orkas
./run.sh           # macOS / Linux
run.cmd            # Windows

run.sh / run.cmd auto-installs dependencies and downloads the embedding model (~95 MB). First launch creates a workspace under ~/.orkas/ (macOS / Linux) or <smallest non-system drive>:\.orkas\ (Windows). Then open Settings → AI Providers to add an API key or OAuth.


Screenshots

Orkas home screen


How it works (core design)

Full design and hard constraints → CLAUDE.md

Group chat: visibility slicing + a single scheduling primitive

In one chat there's a commander, N agents, and you — but each agent does not see the same conversation.

  • Visibility slicing — the main conversation is one full jsonl; each agent only gets a slice (from==me ∨ to∋me ∨ mentions∋me). A worker never reads the full main conversation — saves tokens and prevents private context from leaking across agents.
  • One scheduling primitive — every dispatch (the commander's dispatch_to, the user's @, plan steps) funnels into the same enqueue primitive. No parallel routing paths.
  • Shared plan — when agents collaborate, the commander writes progress into one plan.md, visible to every member.

Agent dispatch: structured channels, not @ in prose

  • Structured dispatch — commander↔agent dispatches go through the dispatch_to({to, message}) tool call; @ in prose is not treated as a dispatch signal (the user's @ is still recognized — UX unchanged).
  • Deferred wake-up — a dispatch_to only stages; the recipient wakes only after the commander's turn finishes, preventing premature execution.
  • Turn-based safety stop — the runaway guard counts turns (MAX_WORKER_TURNS=100), not wall-clock time, so a slow-but-progressing LLM isn't killed.

Self-evolution: meta/ + self-managed skills

Each agent maintains, in its own directory:

  • meta/COMPETENCE.md — what it's good / not good at.
  • meta/LEARNING_STRATEGIES.md — methods that have worked for it.

After each task the agent reflects and updates these; on the next task meta/ is fed back into the system prompt, so experience shapes the next run. Via the skill_manage tool an agent can also crystallize "how I solved X" into a private skill, reused directly next time.


Acknowledgments

Some core modules draw on these open-source projects — special thanks to:


License

MIT

Ver o original no GitHub

Bom para

Relacionados