dev.tokensafe/tokensafe

not probeable (local package)

Unclaimed · Unverified

Publisher-provided title

TokenSafe

Publisher-provided description

Cut AI coding agent token usage up to 90% via AST slicing, dependency graphs & security guards.

Summary by lookupd

Lets agents read code with AST slicing and dependency graphs to reduce tokens, with security guards applied to the code they access.

Install

Claude Code, remote (streamable-http)

claude mcp add --transport http tokensafe https://api.tokensafe.dev/mcp

Claude Code, npm package (stdio)

claude mcp add tokensafe -- npx -y tokensafe

Claude Code, pypi package (stdio)

claude mcp add tokensafe -- uvx tokensafe

Codex, remote (streamable-http)

codex mcp add tokensafe --url https://api.tokensafe.dev/mcp

Codex, npm package (stdio)

codex mcp add tokensafe -- npx -y tokensafe

Codex, pypi package (stdio)

codex mcp add tokensafe -- uvx tokensafe

Cursor, remote (streamable-http) → .cursor/mcp.json

{
  "mcpServers": {
    "tokensafe": {
      "url": "https://api.tokensafe.dev/mcp"
    }
  }
}

Cursor, npm package (stdio) → .cursor/mcp.json

{
  "mcpServers": {
    "tokensafe": {
      "command": "npx",
      "args": [
        "-y",
        "tokensafe"
      ]
    }
  }
}

Cursor, pypi package (stdio) → .cursor/mcp.json

{
  "mcpServers": {
    "tokensafe": {
      "command": "uvx",
      "args": [
        "tokensafe"
      ]
    }
  }
}

Other clients: docs. Or tell your agent: Set up lookupd for me: read https://lookupd.com/agent.md and follow it.

Newest first, as lookupd saw them

Shows this live status and links here; updates hourly.

[![lookupd status](https://lookupd.com/badge/dev.tokensafe/tokensafe.svg)](https://lookupd.com/servers/dev.tokensafe/tokensafe)