AgentDNSAgentDNS
ExploreCompareAPI DocsBlogStatsDashboard
Register Agent
The open registry for AI agents

DNS for the
Agent Economy

Discover AI agents by capability. Resolve endpoints in milliseconds. Trust scores backed by real data.

Try:contract-reviewcode-generationdata-analysisimage-generation
+registered agents

Discover

Find agents by capability, protocol, or category. Search the entire agent ecosystem from one place.

Resolve

Get agent endpoints, capabilities, and communication preferences in a single API call. Built for machine speed.

Trust

Trust scores computed from real usage data. Verified agents, lookup analytics, and reputation tracking.

0+
Agents indexed
0
Data sources
0%
Free API access
0/7
Live monitoring

Protocols & Frameworks

A2A Protocol
MCP Servers
OpenAPI Tools
LangChain Agents
AutoGPT Plugins
CrewAI Agents
Semantic Kernel
Function Calling
Agent Protocol
Tool Use APIs
RAG Pipelines
Webhook Agents
A2A Protocol
MCP Servers
OpenAPI Tools
LangChain Agents
AutoGPT Plugins
CrewAI Agents
Semantic Kernel
Function Calling
Agent Protocol
Tool Use APIs
RAG Pipelines
Webhook Agents
A2A Protocol
MCP Servers
OpenAPI Tools
LangChain Agents
AutoGPT Plugins
CrewAI Agents
Semantic Kernel
Function Calling
Agent Protocol
Tool Use APIs
RAG Pipelines
Webhook Agents
A2A Protocol
MCP Servers
OpenAPI Tools
LangChain Agents
AutoGPT Plugins
CrewAI Agents
Semantic Kernel
Function Calling
Agent Protocol
Tool Use APIs
RAG Pipelines
Webhook Agents

How it works

Three steps to agent resolution

From search to integration in under a minute. Works the same whether you are a developer or an AI agent.

1

Search

Query by capability, category, or name. The API returns matching agents ranked by trust score. Works for humans and machines.

2

Evaluate

Check trust scores, verification status, and endpoint health. Compare agents side by side. Pick the one that fits.

3

Integrate

Get the endpoint URL, protocol details, and authentication requirements. Call the agent directly or resolve at runtime via the API.

Features

Everything you need to find and trust agents

Built for developers and AI systems that need reliable agent discovery at scale.

Agent Discovery

Search thousands of agents by capability, category, or protocol. Full-text search with fuzzy matching finds what you need fast.

Trust Scores

Every agent gets a trust score from 0.10 to 0.99. Computed from endpoint health, metadata quality, usage data, and verification status.

Protocol Support

Works with MCP, A2A, REST, GraphQL, and WebSocket agents. One registry for every protocol. No vendor lock-in.

Compare Agents

Side-by-side comparison of capabilities, trust scores, protocols, and pricing. Pick the right agent for your use case.

API Access

RESTful API with OpenAPI spec. 60 requests/min free, 600 authenticated. Resolve agents programmatically in your pipelines.

Agent Registration

Register your agent in minutes. Sign in with GitHub, fill in the details, and make your agent discoverable to the ecosystem.

For Developers

Built for machine-to-machine discovery

Your agent can discover other agents at runtime. One API call resolves capabilities to endpoints.

resolve.ts
// Resolve agents by capability
const response = await fetch(
"https://agent-dns.tech/api/v1/resolve"
+ "?capability=contract-review"
+ "&protocol=a2a"
);
// Response
{
"matches": [{
"name": "LegalReview AI",
"trust_score": 0.94,
"a2a_endpoint": "https://legal.ai/.well-known/agent",
"capabilities": ["contract-review", "legal-analysis"]
}]
}

OpenAPI specification

Full OpenAPI 3.0 spec at /api/v1/openapi.json. Generate clients in any language automatically.

Runtime resolution

Your AI agent can discover other agents at runtime. No hardcoded endpoints. Dynamic capability resolution.

Standards compliant

Supports /.well-known/agents.json discovery. Compatible with A2A Agent Cards and MCP server manifests.

Read the full API docs

Protocol Support

One registry, every protocol

AgentDNS is protocol-agnostic. Register and discover agents regardless of their communication protocol.

A2A
Native

Agent-to-Agent Protocol

Google's open protocol for AI agent interoperability. Full Agent Card support.

MCP
Native

Model Context Protocol

Anthropic's standard for connecting AI to tools and data. Server manifest indexing.

REST
Supported

RESTful APIs

Standard HTTP endpoints with OpenAPI specs. The most common agent interface today.

GraphQL
Supported

GraphQL APIs

Typed query APIs for complex agent interactions. Schema introspection supported.

Why AgentDNS

The AI agent ecosystem is fragmented. Thousands of agents exist but most operate in isolation. AgentDNS is the missing discovery layer.

Protocol-agnostic

One registry for MCP, A2A, REST, GraphQL, and WebSocket agents. No vendor lock-in.

Trust scores you can verify

Algorithmic scoring based on endpoint health, usage data, metadata completeness, and verification status.

Built for machines

Your agent can discover other agents at runtime. One API call returns endpoints, trust scores, and protocols.

Open and free

9,500+ agents indexed. Free API. No paywall. Register your agent in minutes with GitHub.

Frequently asked questions

Ready to join the agent economy?

Make your agent discoverable to thousands of developers and AI systems. Or explore the registry to find the perfect agent for your use case.

Register your agentExplore the registry
AgentDNSAgentDNS

DNS for the agent economy. Discover, resolve, trust.

Product

ExploreRegister AgentImportStats

Developers

API DocsIntegrationsStandardsAPI Status

Resources

BlogDashboardPrivacy PolicyTerms of Service
AgentDNS
PrivacyTermsThe open registry for AI agents