Index / Verified tools

Verified MCP tools, by capability

Every other registry indexes servers: a name, a URL and whatever prose the publisher wrote. The tool names and input schemas an agent actually has to match on exist in no index. So we ask each server directly, and store what it answers.

Maps, location and weather
11,222 tools · 4,588 servers
Analytics and monitoring
10,657 tools · 4,896 servers
Calendar and scheduling
9,134 tools · 2,238 servers
Legal and government data
8,661 tools · 1,376 servers
Files and storage
8,395 tools · 4,911 servers
Science and research
8,341 tools · 3,979 servers
Crypto and blockchain
6,869 tools · 2,042 servers
GitHub and version control
6,255 tools · 3,824 servers
Payments and billing
6,166 tools · 3,316 servers
Web scraping and browser automation
5,555 tools · 2,588 servers
AI models and embeddings
3,967 tools · 1,473 servers
Images, audio and video
2,738 tools · 583 servers
Project management
2,548 tools · 2,342 servers
HR and recruiting
2,033 tools · 569 servers
E-commerce
1,912 tools · 547 servers
PDF and document extraction
1,880 tools · 583 servers
Cloud and DevOps
1,574 tools · 280 servers
Finance and market data
1,507 tools · 552 servers
Email
1,251 tools · 454 servers
Social media
1,232 tools · 173 servers
Databases and SQL
1,003 tools · 222 servers
CRM and sales
650 tools · 281 servers
Security and compliance
584 tools · 230 servers
Web search
466 tools · 171 servers
Messaging and chat
462 tools · 149 servers
Translation and language
195 tools · 84 servers

What is in the index

Measured, not claimed. This is the whole index by kind, 124,897 resources in total.

MCP servers116,125
Agent skills4,342
OpenAPI services2,883
Other callable resources716
Machine-readable docs538
A2A agents137
Plugin manifests107
Catalogues21
ARD registries16
Packages7
GraphQL APIs2
Agent descriptors (ACP, OASF, AgentFacts)1
Datasets1
WebMCP browser tools1

How the ecosystem spells the same thing

Media types as publishers actually wrote them. Filters match exactly, so a registry that does not normalise these silently drops thousands of entries. We normalise all of them, plus both URN prefixes in circulation (urn:air: and urn:ai:), on ingest.

application/mcp-server-card+json114,513
application/ai-skill3,024
application/vnd.oai.openapi+json2,614
application/ai-skill+md1,227
application/mcp-server1,072
application/json545
application/mcp-server+json533
application/openapi+json214
text/markdown207
text/html144
application/a2a-agent-card+json118
application/vnd.anthropic.claude-plugin+json83
This is not a directory of every server. We measured what ranks for MCP queries and the primary source wins: mirroring it at scale adds nothing. These pages exist because the verified tool surface is data that is not published anywhere else.

The whole corpus is also an open dataset under CC BY 4.0: huggingface.co/datasets/AgenticResourceDiscovery/verified-mcp-tools.