Orchestration tools for running multiple Hermes agents together. Fleet management dashboards, swarm coordinators, and multi-agent delegation systems.
A self-hosted orchestration layer that manages multi-agent workflows across CrewAI, LangGraph, and AutoGen. It features a zero-dependency SQLite stack and an Aegis review system to enforce quality gates on dispatched tasks.
A TUI and PWA dashboard designed to monitor and manage various coding agents, including Hermes Agent and Claude Code. It ensures execution isolation using Docker or Podman and enables parallel work via git worktrees.
A governance-focused template that provides a sidecar control plane for managing everything from single VPS agents to specialist teams. It utilizes task bus patterns to structure delegation between orchestrators and specialists.
A framework for building autonomous agent swarms with support for over 23 LLM providers, including Ollama and OpenRouter. It includes visual org charts for team orchestration alongside durable memory and scheduled execution.
A scaling template that uses a level-based architecture to transition from single agents to automated teams. It bundles specific skills for security auditing and VPS provisioning via a task bus delegation pattern.
A supervision interface for Hermes agents that uses a Kanban board to track task status and review queues. It provides live streaming of agent reasoning and requires human-in-the-loop verification for all task completions.
A local collaboration system that orchestrates Leader and Specialist agents using MCP and ACP protocols. It features a real-time Web UI to monitor agent dialogues and Kanban task flows.
A specialized collective of 17 agents dedicated to research, planning, and infrastructure. The framework enforces mandatory quality gates through dedicated reviewer and tester agents with priority-based conflict resolution.
A meta-optimization harness that uses automated outer-loop search to refine agent code. It orchestrates TBLite and TB2 benchmark evaluations to track performance frontiers via candidate mutation.
An agent-vs-agent arena where autonomous AI companies compete for GitHub stars. It orchestrates nine agents using Paperclip and Hermes Agent, visualizing the competition through a real-time scoreboard and task dashboard.
An ACP-style delegation system that routes tasks between Hermes subagents, Codex, and Claude Code. It prevents parent context pollution through isolated workflows and enforces safety via character limits and timeouts.