Skip to content
View Orkas-AI's full-sized avatar

Block or report Orkas-AI

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Orkas-AI/README.md

Orkas — an open-source, local-first multi-agent AI desktop app for commanding a team of AI agents

Orkas-AI

Orkas-AI builds Orkas — an open-source, local-first desktop app for directing a team of AI agents with your own LLM keys.

In Orkas you direct a commander agent by chat. Beyond routing work, the commander brings the strengths coding agents are known for — precise file read/write, careful tool operation, engineering discipline, multi-step reasoning, and long-horizon execution — and applies them across a wide range of tasks, not just coding. When a job is bigger than one agent, the same commander assembles specialized sub-agents and runs them in parallel or in sequence — no flowcharts, just conversation.


What is Orkas?

Orkas is not a local LLM runtime and not a hosted SaaS. It is a local-first desktop AI collaboration client where a commander agent both does the work itself and coordinates specialized sub-agents — across coding, research, content creation, data analysis, video, slides, and office automation. Your files, chats, and LLM keys stay on your machine, and model traffic is never proxied through Orkas.

Why Orkas?

  • A capable commander, not just a router — the lead agent brings coding-agent strengths (precise file I/O, careful tool use, engineering discipline, multi-step reasoning, long-horizon execution) to many kinds of tasks
  • Multi-agent orchestration by chat — one commander assembles many specialized sub-agents, in parallel or in series; conversation is the orchestration, no flowcharts required
  • Drives the open-source ecosystem — command external CLI coding agents (Claude Code, Codex, OpenCode, Cline) and onboard other open-source projects (e.g. HyperFrames) as local packages, so one commander can deliver code, video, slides, and more
  • Local-first — conversations, files, and LLM keys stay on your machine; model traffic is never proxied
  • Bring your own keys — Claude, OpenAI, Gemini, DeepSeek, Qwen, Kimi, GLM, and more
  • Reusable skills — turn repeated workflows into skills and sub-agents your team can call again
  • Self-evolving agents — agents review and evolve their own approach through memory and reflection
  • Open source — MIT licensed, for macOS / Windows / Linux

Repositories

  • Orkas — the main open-source multi-agent AI desktop app
  • Orkas-Awesome-AgentSkills — a curated library of agents and reusable skills for education, product development, content creation, data analysis, and office automation

Compared with

Orkas differs from LangChain, CrewAI, Claude Desktop, and n8n: it is a no-code desktop app where a capable commander agent does the work with the precision, tool use, and engineering discipline of a coding agent — and commands an AI agent team, and can drive open-source projects (from CLI coding agents like Claude Code to tools like HyperFrames) to cover many task types — all while keeping your data and LLM keys local on your own machine.


Direct your AI agent team by chat — get Orkas at orkas.ai.

Popular repositories Loading

  1. Orkas Orkas Public

    Open-source multi-agent AI desktop client — build and command your AI agent team through conversation. A commander LLM dispatches sub-agents in parallel or in series; agents self-evolve via reflect…

    TypeScript 32 8

  2. Orkas-Awesome-AgentSkills Orkas-Awesome-AgentSkills Public

    Curated Orkas agents and skills for education, product development, content creation, data analysis, office automation, and e-commerce workflows.

    Python 10 4

  3. Orkas-AI Orkas-AI Public