Productivity

Capture and replay every AI agent tool call, prompt, and response to trace sessions and debug complex workflows with full observability.

Orchestrate multi-agent workflows with an SDK-first framework offering node types, parallel execution, retries, and cost tracking for MCP integrations.

Background-capable Claude Code wrapper with async workflows, session management, and multi-host orchestration. Telegram Integration, TTS-STT with multi-LLM transcript-to-structured-prompt pipeline

Github Agentic Workflows MCP Gateway

Open-source omnichannel chatbot for agentic workflows via APIs, CLI, and MCP. An alternative to Wati, ManyChat, and Respond.io 🤖 🚀

An operating contract, workflow discipline, and knowledge propagation harness for orchestrating AI agents across any number of repositories. One hub, one contract, N repos, shared knowledge flows both ways.

Apple-first control surface for Codex-powered agent workflows with a Mac companion and relay service.

Catch silent failures in agent workflows by verifying real system state

BUDDY.OS Marketplace - Buy and Sell AI Agents, Workflows & Apps

Self-hosted voice AI — TTS, STT, voice cloning, conversational agents, workflows. Independent of ElevenLabs and n8n.

An layer for agentic tools / workflows , to interact with local llm ( nemotron-3-nano:4b )

Run a local HUD-style AI desktop assistant for chatting, commands, agent workflows, and fast PC control with cloud or local mode

Run an always-on Claude Code AI assistant for nonstop coding help and workflow automation

Multi-agent workflow system for Claude Code

Orchestrate multi-agent workflows using a structured, centralized decision system without Python or dependencies for Claude Code integration.

Reusable MCP-driven E2E harness for agent workflows

Agent-human interaction protocol for agentic workflows. Human-in-the-loop primitives for AI agents.

Successor to ccg-workflow. Multi-model dev workflow for Claude Code — auto-routes frontend → Gemini, backend → Codex, with quality tiers, autonomous long-runs.

Conductor is an event driven agentic workflow engine providing durable and highly resilient execution engine for applications and AI Agents

Pytest-based black-box validation framework for an external Fraud Detection MCP servers, covering MCP tool correctness, agentic workflow behaviour, security boundaries, observability, and regression evidence.

Sovereign, local-first secrets manager for AI agent workflows. age-encrypted at rest, referenced by handle in chat. No daemon, no cloud.

🚀 Turn your LLM into an Scrum team. An E2E agent workflow using Todoist MCP to architect, batch, execute, and verify code via structured state machines.

Python CLI library for AI agents to interact with Google Calendar, Gmail, and Notion APIs

A complete Agentic AI repository covering AI agents, memory systems, reasoning, multi-agent workflows, and tool integrations using modern AI frameworks.