ai-orchestrator

Windows desktop app with an adaptive AI orchestrator for Flowseal strategies, automatic tuning and network-aware profile selection

Vigilante is a sandbox-first orchestration layer for coding agents. It isolates every task in a git worktree, enforces strict credential scoping, and gives you full audit logs — so your agents can't burn down production.

A high-performance, open-source Multi-Agent Collaboration Platform built with Tauri, Rust, and SolidJS

A lightweight orchestration framework that piggybacks your local Agentic CLI setup. Intentionally simple. Yet powerful... like an army of Orcs 👹

Cloud-independent, self-hosted AI agent runtime — no vendor lock-in. Slack · Telegram · Discord · Web. 9 provider-neutral backends (Claude, Codex, Gemini, OpenAI, Ollama…), 141-node workflow engine, multi-agent Phase Loops, HITL gates, AES-256-GCM Vault. TypeScript. AGPL-3.0.

A structured orchestrator for AI coding agents. Give it a request — it plans, implements, and reviews the work automatically using your preferred coding CLI.

Async AI coding orchestrator: turns Yandex Tracker tasks into Claude Agent SDK workflows and GitHub PRs, with code review, deployment checks, and supervised auto-merge.

Fugusashi — Federated Learning of LLM Routing with Human-Interpretable Decisions. Open-source alternative to Sakana AI Fugu.

Discoclaw: CLI-first Discord bridge for agent runtimes

🔬 Parallel AI research orchestration tool - Deploy multiple cursor-agent workers simultaneously to conduct comprehensive research with specialized focus areas and intelligent synthesis.

Atomic Spec: AI-native agentic methodology for living requirements. DDD + TDD + Use-Case-Driven. One atom, one file, one unit of knowledge.

Self-hosted AI orchestrator — 100+ MCP tools, 6-provider cascade, RAG hybrid. Automate the chaos on your own server. An Automatia BCN product.

Multi-provider AI router with cost tracking, RAG memory and live dashboard. Routes tasks across Claude, OpenAI, DeepSeek and Gemini via CLI and MCP server.

A workbench for agentic development: model catalog, workspaces of dedicated agents behind one orchestrator chat, blueprint wiring, and a persistent catalog of agent-forged tools. Local-first, no telemetry.

Control two agents working together (in a head-to-head format) to complete advanced user tasks.

Fully local, modular, event-driven multimodal AI assistant with voice input, local LLM reasoning, Piper TTS, barge-in support, performance metrics, and future UI/avatar/screen vision capabilities.

An autonomous agent orchestration wrapper for Windows. Features a native Electron desktop UI, Python FastAPI backend, vision LLM integrations, and fully local audio transcription.

Run Claude Code as orchestrator, Codex CLI as worker. The playbook for the parts the official docs leave to you.

Agent Orchestration System - TypeScript framework for building autonomous, collaborative AI agents

Local multi-provider AI task runner with provider fallback, workflow loop, git automation, and project memory.