The trusted catalog and delivery path for MCP servers, agents, and skills.

---

Agent Registry gives platform teams and developers one place to manage the agentic infrastructure their applications depend on.

Use the web UI and arctl CLI to publish approved MCP servers, agents, and skills, discover what is available, and make those artifacts usable across local development, shared environments, and Kubernetes.

---

🤔 Why Agent Registry?

• 📦 One trusted source for AI building blocks — a curated catalog instead of scattered repos, scripts, and one-off MCP setup.
• 🚀 Faster developer onboarding — discover approved artifacts quickly with less manual configuration.
• 🌐 Consistent path from laptop to cluster — same discovery and delivery workflow across local dev and Kubernetes.
• 🔒 Governance without slowing teams down — centralize curation and publishing without forcing each team to rebuild the process.

🔗 Quick Links

• 📥 Install arctl
• 🚀 Quickstart guides:
  • Local development
  • Kubernetes
• 🎬 See it in action
• 📖 Development details
• 🤝 Contributing
• 💬 Discord

---

📚 Core Capabilities

📦 Registry

Curate a shared catalog of MCP servers, agents, and skills your teams can trust and reuse.

• Publish artifacts from a central registry
• Discover approved artifacts with the CLI and web UI
• Give teams a consistent source of truth across environments

🔒 Curation and Governance

Turn a broad set of available AI artifacts into a collection your organization is willing to support.

• Organize what developers can discover and deploy
• Standardize how artifacts are shared across teams
• Keep control of what gets published and promoted

🚀 Deployment Workflows

Move from discovery to usage without reinventing the same delivery path for every team.

• Run workflows locally with arctl
• Deploy Agent Registry into Kubernetes with Helm
• Support local environments and shared platform environments from the same registry

🌐 Client and Gateway Integration

Make approved artifacts easier to consume from the tools developers already use.

• Generate configuration for Claude Desktop, Cursor, and VS Code
• Pair with Agent Gateway for a consistent access layer to deployed MCP infrastructure
• Reduce manual setup for AI clients and shared environments

🔧 How It Works Together

1. Platform teams curate and publish approved MCP servers, agents, and skills in Agent Registry.
2. Developers discover those artifacts through the web UI or arctl.
3. Teams pull and deploy what they need in local environments or Kubernetes.
4. AI clients and shared gateway infrastructure connect to approved artifacts through a consistent workflow.

---

🏗️ Flexible Deployment

💻 Local Development

Get started with a local registry in minutes. The first time arctl runs, it automatically starts the local registry daemon and imports the built-in seed data.

# Install via script
curl -fsSL https://raw.githubusercontent.com/agentregistry-dev/agentregistry/main/scripts/get-arctl | bash

# Discover available MCP servers
arctl mcp list

# Configure supported AI clients
arctl configure claude-desktop
arctl configure cursor
arctl configure vscode
arctl configure claude-code

Open http://localhost:12121 to use the web UI.

☸️ Kubernetes

Run Agent Registry in a cluster when you want shared discovery and deployment workflows. An external PostgreSQL instance with the pgvector extension is required.

PostgreSQL

Deploy a single-instance PostgreSQL and pgvector into your cluster using the provided example manifest:

kubectl apply -f https://raw.githubusercontent.com/agentregistry-dev/agentregistry/main/examples/postgres-pgvector.yaml
kubectl -n agentregistry wait --for=condition=ready pod -l app=postgres-pgvector --timeout=120s

This setup is intended for development and testing. For production, use a managed PostgreSQL service or a production-grade operator.

Install Agent Registry

helm install agentregistry oci://ghcr.io/agentregistry-dev/agentregistry/charts/agentregistry \
  --namespace agentregistry \
  --create-namespace \
  --set database.host=postgres-pgvector.agentregistry.svc.cluster.local \
  --set database.password=agentregistry \
  --set database.sslMode=disable \
  --set config.jwtPrivateKey=$(openssl rand -hex 32)

Then port-forward to access the UI:

kubectl port-forward -n agentregistry svc/agentregistry 12121:12121

Get started: Helm chart details, Local Kind cluster

---

🎬 See It In Action

Learn how to create an Anthropic Skill, publish it to Agent Registry, and use it in Claude Code.

---

🤝 Contributing

We welcome contributions and feedback from the community!

• 🐛 Report issues
• 💡 Start a discussion
• 🔧 Contributing guide
• 📖 Development details
• 💬 Join our Discord

---

📄 License

Apache V2 License. See LICENSE for details.