#2020th of 62 in Agent and MCP starters

agenticx

Python multi-agent framework with an Electron desktop client and Go gateway

Stars
1k
License
Apache-2.0
Last commit
Nov 2026
Last release
Sep 2026
Language
Python
Stars, last 29 weeks+929
22 Apr 2026: 115 stars29 Apr 2026: 117 stars6 May 2026: 118 stars13 May 2026: 120 stars20 May 2026: 122 stars27 May 2026: 125 stars3 Jun 2026: 134 stars10 Jun 2026: 152 stars17 Jun 2026: 163 stars24 Jun 2026: 166 stars1 Jul 2026: 177 stars8 Jul 2026: 181 stars15 Jul 2026: 184 stars22 Jul 2026: 190 stars29 Jul 2026: 201 stars5 Aug 2026: 206 stars12 Aug 2026: 210 stars19 Aug 2026: 213 stars26 Aug 2026: 218 stars2 Sept 2026: 227 stars9 Sept 2026: 230 stars16 Sept 2026: 235 stars23 Sept 2026: 235 stars30 Sept 2026: 240 stars7 Oct 2026: 295 stars14 Oct 2026: 424 stars21 Oct 2026: 587 stars28 Oct 2026: 774 stars4 Nov 2026: 969 stars11 Nov 2026: 1,044 stars

115 on 22 Apr 20261,044 on 11 Nov 2026

From GitHub's daily star counts.

Checked on GitHub.

Overview

AgenticX is a Python agent SDK and `agx` CLI with a FastAPI/SSE Studio server, covering a ReAct loop, graph workflows, MCP and A2A, memory, knowledge bases and sandboxed code execution. It also ships Near Desktop, an Electron app that talks to a local `agx serve`, and an Enterprise edition with a Go AI Gateway for auth, quota, audit and model relay.

Who it is for: Python developers building multi-agent apps, optionally with desktop or gateway

Strengths

  • Core install is about 27 lightweight dependencies; torch and others are optional extras
  • 15+ LLM providers including OpenAI, Anthropic, Ollama, Gemini, with failover routing
  • Supports MCP, A2A, AG-UI and OpenAPI tools, plus Docker, Microsandbox and subprocess sandboxes
  • Ships a desktop client and an enterprise gateway with audit, not just a library

Weaknesses

  • Very broad scope; some storage adapters are placeholders or not wired into Studio
  • Studio path does not yet route every operation through the full SafetyLayer
  • Enterprise Cluster Agent Runtime is only planned; Edge Agent is an MVP
  • Near Desktop and Enterprise use independent runtime paths today

What it needs

  • Python, OpenAI, Anthropic, Ollama, Gemini
  • Needs Python 3.10+, Redis (optional), PostgreSQL or MySQL (Enterprise), Chroma (Desktop knowledge base default), Docker or Microsandbox (optional sandbox)
  • Docker

What changed

From the release notes. No flag means the note does not say so, not that the upgrade is safe.

Also in Agent and MCP starters

See all 62
Also in Agent and MCP starters
RankProjectScore
1deepseek-harnessPlugin-based agent harness from DeepSeek with a local Web UI279.6k stars+32,096 (+13%) in 30 daysMIT98 out of 100
2composioTypeScript and Python SDKs that give agents authenticated tools across 1000+ apps31k stars+574 (+1.9%) in 30 daysMIT92 out of 100
3conductorDurable workflow engine that orchestrates microservices and LLM agents from JSON32.5k starsApache-2.091 out of 100
#4a2uiProtocol and renderers for agents to stream native UI instead of text Live demo ↗ (opens in a new tab)17.2k starsApache-2.087 out of 100
#5mcp-useTypeScript framework for building MCP servers with React Views and an Inspector Live demo ↗ (opens in a new tab)10.9k stars+132 (+1.2%) in 30 daysMIT87 out of 100