# Metatake MCP server

Source: https://glimind.com/server/mcp-registry%2Fnet.metatake%2Fmcp
Registry id: `mcp-registry/net.metatake/mcp`
Status: live (checked 2026-10-26T23:40:27.281Z)

## Verdict

Metatake MCP server answered 99.95% of Glimind's liveness checks over 30 days, better than 36% of MCP servers, with 963 ms median handshake and 33 outages in 90 days. Its agent-readiness grade is A (96/100) across 6 tools, costing an agent about 1,001 tokens to load. Checked every few minutes; no tool is ever called.

## Measurements (30 days)

| Metric | Value | Better than |
|---|---|---|
| Uptime, 30 days | 99.95% | 36% of MCP servers |
| Median handshake latency | 963 ms | 22% of MCP servers |
| p95 handshake latency | 3665 ms | — |
| Checks in 30 days | 4,949 | — |
| Outages, 90 days | 33 | — |
| Median recovery | 5 min | — |
| Agent-readiness score | 96/100 (A) | 82% of MCP servers |
| Tools | 6 | — |
| Tool-list size for an agent | ~1,001 tokens | 73% of MCP servers |
| Days observed | 29 | — |

## Install

```json
{
  "mcpServers": {
    "metatake-mcp-server": {
      "type": "http",
      "url": "https://metatake.net/api/mcp"
    }
  }
}
```

Claude Code:

```sh
claude mcp add --transport http metatake-mcp-server https://metatake.net/api/mcp
```

## Tools (6)

- `fetch`: Retrieve the full text of one Metatake document by its `id` (a film slug from `search`). Returns the film's complete critical pack as text — multi-framework rea
- `find_connected_films`: Kindred films — titles that share interpretive threads (concepts, figures, frameworks) with the given film on Metatake's connection map. Good for 'what should I
- `get_film_criticism`: Metatake's full context pack for one film, as Markdown: critical readings (each pushes one framework as far as the film allows), the 13-dimension TakeScore, can
- `get_takescore`: Metatake's 13-dimension critical assessment for one film: net Value score plus the Value / Cost / Risk dimension breakdown. Available for the whole scored corpu
- `search`: Search Metatake's film-criticism corpus and return a list of matching documents (one per film). Each result has an `id` (the film slug), a `title`, and a `url`.
- `search_films`: Find films in Metatake's corpus by title (also matches original titles and director names). Returns up to 10 matches with year, director, TakeScore and the slug

Measured by Glimind (https://glimind.com): MCP handshake + tool list only, no tool is ever called. CC BY 4.0.