Skip to content

nanotoll MCP server

Which MCP servers are nanotoll's own, from the official registry and from asking its own domain, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.

read 2026-10-26

← Back to nanotoll

Whose server it is
  • Registry namedev.nanotoll/nanotoll-mcpverified in the MCP registry

The registry name sits under nanotoll.dev. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.

When we asked itchecked 2026-10-26
  • Stateanswerschecked 2026-10-26
  • Tools1
  • Server calls itselfrmcp
  • Server version3.4.0
  • Protocol version2025-06-18

On 2026-10-26 the server answered initialize and tools/list without a sign-in. First answer in 654 ms.

Should an agent connect itNot proven either way yet
  • Its tools do every use case we checked: shrink long agent context with extractive token pruning; check compression fidelity through per-call signed receipts.
  • Checked on 1 of the 7 days it takes to judge reliability.

Once connected

  • For "shrink long agent context with extractive token pruning", the likely first call is compress.
  • For "check compression fidelity through per-call signed receipts", the likely first call is compress.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

Measured dailydaily checks: 1, the first on 2026-10-26
  • INITE MCP scoreafter a weeka score needs 7 daily checks: 1 so far
  • What an agent can do with it2 of 2 use cases
  • Answeredmeasuring1 of the 7 daily checks it takes to print a share
  • Median first answer673 ms
  • Space in an agent's context≈ 358 tokensthe definitions of all its tools (1), which an agent carries on every turn
  • Tool definitionsA · 100 of 100
  • Typed results0 of 1 toolsdeclared outputSchema: results come typed, read once
  • Every tool passes all six checks.

What an agent can do with it

  • not sure · strip comments and prune JSON before model calls
  • can do · shrink long agent context with extractive token pruning
  • can do · check compression fidelity through per-call signed receipts

Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?

Worth a look

None of the five warning signs in these definitions.

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

Tools it offers1 tools · read 2026-10-26
  • compress

    Compress text before it reaches an LLM context window. Returns the compressed text plus a signed receipt (ratios, refusals, coverage) so the caller can verify what was dropped. Flat per-call price by mode: deterministic free, extractive…

How to connect

Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:

{
  "mcpServers": {
    "nanotoll-mcp": {
      "type": "http",
      "url": "https://mcp.nanotoll.dev/mcp"
    }
  }
}

Add it

Claude Code, in a terminal

claude mcp add --transport http nanotoll https://mcp.nanotoll.dev/mcp

Add to CursorAdd to VS Code

claude.ai, ChatGPT and other clients: add https://mcp.nanotoll.dev/mcp as a custom connector in the settings.

Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.