Skip to content
Migrate your GPTCheck my InvokeRank
InvokeIndex

The tools agents can call

161,537 unique MCP servers across ten major directories and registries, each with its sources and history.

Each server counts once, however many directories list it. Glama, PulseMCP and mcp.so listings arrive through MCP Toplist’s open dataset, used under CC BY 4.0. 14,492 listings from bulk publishers wait for review.

Check my InvokeRankUpdated How we collect
  • Directory listings65,856Collected from seven directories, before grouping
  • Listed in two or more34,63721% of unique servers
  • Remote MCP servers30,213With a public endpoint an agent can call
  • Changes, last 30 days117,515Typed diffs between observations

Glama (102,362)

Plus 330 listings grouped under these serversPage 455 of 4,266
Inside the index

Where the servers are listed

Each directory is a shelf. See how many shelves each server sits on, how large each shelf is, and what moved this month.

Updated

How many shelves

Servers by the number of directories listing them

79%of servers are listed in one directory only. 34,637 are listed in two or more.

Directories listing the server. Copies grouped under a server are not counted again.

The shelves

Unique servers listed in each directory
  1. Glama102,362
  2. MCP Registry42,804
  3. PulseMCP22,399
  4. mcp.so19,070
  5. Smithery13,426
  6. ChatGPT7,104
  7. Claude6,349
  8. GitHub483
  9. Cursor456
  10. Claude Code336

Collected by us Via MCP Toplist’s dataset

A server listed in several directories counts once in each. Pick a shelf to browse it.

What changed

Typed diffs between observations, last 30 days
New listings 14,589Left a directory 664Tool changes 85,325Listing details 16,937The busiest day was 15 Oct, with 10,748 changes.
Changes by day, last 30 days (days with changes)
DayNew listingsLeft a directoryTool changesListing details
2 Oct26540140
3 Oct647203212969
4 Oct51463265441
5 Oct942293912615
6 Oct6461559201698
7 Oct7372250891699
8 Oct7042415021864
9 Oct16671759321572
10 Oct13976049091137
11 Oct5272150281634
12 Oct7532044771770
13 Oct871215291627
14 Oct853284194665
15 Oct1276208675777
16 Oct11111038540842
17 Oct1066175225830
18 Oct613206635657

Latest changes

RSS feed
  1. Cheapest LLM API Official MCP Registry benchmark_results: input schema changed
  2. Nimbus BCI Official MCP Registry stats.grouped: input schema changed
  3. Nimbus BCI Official MCP Registry stats.grouped: description changed (+"executions. REQUIREMENT (read first): each execution must have been run with an evaluation plan that STORED per-group (per-subject) accuracies for at least 3 groups — e.g. leave-one-subject-out (LOSO) or grouped CV over >= 3 subjects. A plain single-subject CV run (trials split within one subject, one group) stores no group stats and CANNOT be analysed here. If none of the given executions qualifies, the backend fails with ``nimbus.stats.insufficient_groups`` — that error means exactly this: re-run with a multi-subject / group-wise evaluation plan (>= 3 groups), or pass runs from such a sweep. What it does:" -"executions:")
  4. Nimbus BCI Official MCP Registry protocol.start_assignment: input schema changed (-name)
  5. Nimbus BCI Official MCP Registry protocol.save: input schema changed (-name)
  6. Nimbus BCI Official MCP Registry project.delete: tool added
  7. Nimbus BCI Official MCP Registry plots.confusion: input schema changed
  8. Nimbus BCI Official MCP Registry pipeline.validate_node: input schema changed (+dataset)

Coverage

What the index resolves and probes
Brands resolved
29,034
Brands on two or more hosts
6,725
Held for review
14,492 listings
Tools listed by probe
12,345 servers

Where are you listed?

Paste your MCP server, plugin or company URL to see your listings, tools and scan findings in one place.

Check my InvokeRank Scan a server