~ / directory / agentregistry
AG
Mixed · AI Development Tools · reviewed 2026-04

agentregistry

Open-source centralised registry for AI agents, MCP servers, and skills. Build, publish, discover, and govern AI artifacts with CLI and web UI. Pairs with Agent Gateway. By Solo.io.

Visit aregistry.ai
01

What it does

Open-source platform that provides a single place to find, manage, and run MCP servers, AI agents, and skills. Solves the problem of AI artifacts being scattered across npm, PyPI, Docker Hub, GitHub repos, and random URLs with no trust signals. Features include: centralised catalog with CLI (arctl) and web UI, one-command deployment of agents and MCP servers, automatic IDE configuration for Claude Desktop/Cursor/VS Code, semantic search powered by pgvector, and automated scoring and validation of imported artifacts. Pairs with Agent Gateway (agentgateway.dev) to provide a single secure entry point to all deployed MCP servers and agents. By Solo.io, with contributions from Microsoft, Apple, Alibaba, Adobe, AWS, Cisco, Salesforce, Huawei, and Amdocs.

02

Security relevance

Addresses the supply chain security gap in the AI agent ecosystem. Platform teams curate and publish approved MCP servers and agents; developers discover and deploy from the vetted catalog instead of pulling from unverified GitHub repos. Centralised governance means every artifact goes through approval before company-wide deployment. When paired with Agent Gateway, provides authentication and authorisation before requests reach MCP servers, plus full audit trails of agent-to-tool communication.

03

When to use it

Use when your organisation has multiple teams deploying MCP servers and agents and you need a trusted internal registry — similar to how you'd use a private npm or Docker registry but purpose-built for AI artifacts. Runs locally via Docker Compose or in Kubernetes. CLI-first workflow. Best suited for platform engineering teams that want to standardise how AI tools are discovered, approved, and deployed across the organisation.

04

OWASP coverage

Risks addressed — mapped to both OWASP Top 10 standards. 1 in LLM, 3 in Agentic.

LLM Top 10 · 2025 · 1/10 covered
01
02
03
04
05
06
07
08
09
10
05

The raw record

What Yuntona stores. Single source of truth — fork it on GitHub.

name: agentregistry
slug: agentregistry
type: Mixed
category: AI Development Tools
url: https://aregistry.ai

reviewed:   2026-04
added:      2026-04
updated:    2026-04

risks:
  llm:  [LLM03]
  asi:  [ASI02, ASI04, ASI08]

complexity:    Plug & Play
pricing:       —
audience:      Builder
lifecycle:     [build]

tags: [Agent Security, DevOps, MCP Security, Open Source, Registry]