archestra-ai/archestra

GitHub

2,700

MCP-native Secure AI Platform - Enterprise-grade OpenClaw/ClawdBot/MoldBot for organizations

特性

MCP registry & orchestrator

Kubernetes-native deployment

Security sub-agents (Dual LLM)

Non-probabilistic security to prevent data exfiltration

Cost monitoring and dynamic optimization (up to 96% reduction)

Private MCP registry with governance

ChatGPT-like chat with MCPs

Observability (metrics, traces, logs)

Lightning fast (45ms at 95p)

Terraform provider & Helm Chart

对比

Enterprise-focused with security, cost optimization, and orchestration. Ideal for organizations needing centralized MCP management, security controls, and multi-tenant deployments.

使用场景

Platform teams managing AI adoption, developers deploying agents org-wide, management seeking cost control and visibility

在 GitHub 上查看

nanobot-ai/nanobot

GitHub

1000+

Build MCP Agents - Standalone, open-source MCP host for creating dedicated MCP and MCP-UI powered chatbots

特性

Flexible MCP host

Single file or directory-based configuration

OpenAI and Anthropic provider support

Web UI included

Easy deployment via Homebrew

Agent directory structure

MCP server integration

Examples: Blackjack, Hugging Face, Shopping/Shopify

对比

Lightweight MCP host focused on simplicity. 99% less code than full OpenClaw. Great for dedicated chatbot deployments and MCP experimentation.

使用场景

Developers building custom MCP-powered chatbots, embedded applications, or learning MCP architecture

在 GitHub 上查看

开源项目

archestra-ai/archestra

特性

对比

使用场景

nanobot-ai/nanobot

特性

对比

使用场景