MCPLow riskUnclaimed

cap-shield

Context selection and compression for AI agents - recall measured, not claimed.

robinlidberg-dotrobinlidberg-dot/cap-shieldUpdated Aug 26, 2026

server.json

{
  "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
  "name": "io.github.robinlidberg-dot/cap-shield",
  "description": "Context selection and compression for AI agents - recall measured, not claimed.",
  "repository": {
    "url": "https://github.com/robinlidberg-dot/cap-shield-mcp",
    "source": "github"
  },
  "version": "0.4.1",
  "websiteUrl": "https://cap-shield-robin.fly.dev",
  "packages": [
    {
      "registryType": "pypi",
      "identifier": "cap-shield",
      "version": "0.4.1",
      "transport": {
        "type": "stdio"
      },
      "environmentVariables": [
        {
          "description": "Only needed for the two tools that store and retrieve memory. measure_traffic and list_packages work without it.",
          "format": "string",
          "isSecret": true,
          "name": "CAP_SHIELD_API_KEY"
        }
      ]
    }
  ]
}

Permissions

DeclaredDetected
Runs code—python
Installs—pypi:cap-shield@0.4.1
Runs install scripts—None
NetworkNoneNone
Needs credentialsCAP_SHIELD_API_KEYCAP_SHIELD_API_KEY
Outside the workspace—None
Agent tools—None

Checks

Low risk · Nothing worth a warning was found.

Not reviewed by a person · Checked by rules; the model review is not switched on yet.

Versions

  1. #10.4.1latestOct 7, 2026