Skip to content

Agent Permissions Index

by agent-permissions-index.vercel.app · remote · no sign-in to list tools · v1.0.0 · published by app.vercel.agent-permissions-index

Grade Claude Code, Gemini CLI and Cursor configs for risky permissions; get public config stats.

Changes · 11 high finding
CompareRegistry entry

3 tools: 1 change things and 2 only read.

Scanned Oct 21, 2026

1 check matched something in the tool definitions. No finding is not a safety guarantee. How the checks work.

claude mcp add --transport http agent-permissions-index https://agent-permissions-index.vercel.app/mcp

Findings

Matches of fixed checks, with the text they matched.

1 high
  • Highgrade_config

    The description refers to "mcp.json".

    Matched text

    description: "...n, Gemini CLI .gemini/settings.json, or mcp.json). Returns a letter grade, findings, a s..."

    What this check looks for

Tools

Open a tool for its reasons and full description. Descriptions are the publisher's own words, shown as they are.

3 tools

Changes things 1Claude Code: ask
  • grade_configGrade an AI coding agent permissions config (Claude Code .claude/settings.json, Gemini CLI .gemini/settings.json, or mcp.json).1 finding

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.

    Findings

    • High

      The description refers to "mcp.json".

      Matched text

      description: "...n, Gemini CLI .gemini/settings.json, or mcp.json). Returns a letter grade, findings, a s..."

      What this check looks for
    Description, as the server wrote it
    Grade an AI coding agent permissions config (Claude Code .claude/settings.json, Gemini CLI .gemini/settings.json, or mcp.json). Returns a letter grade, findings, a safer config and README badge Markdown. Inputs are not stored or logged.
Only reads 2Claude Code: allow
  • get_index_statsHeadline aggregate numbers from the latest weekly snapshot of public AI agent configs on GitHub.No findings

    Why it is in this group

    • The name "get_index_stats" says "get": a word for reading.
    Description, as the server wrote it
    Headline aggregate numbers from the latest weekly snapshot of public AI agent configs on GitHub.
  • get_leaderboardAgent tools ranked by how safely their public configs are set up, with sample sizes and method.No findings

    Why it is in this group

    • The name "get_leaderboard" says "get": a word for reading.
    Description, as the server wrote it
    Agent tools ranked by how safely their public configs are set up, with sample sizes and method.

This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.

Install

Pinned to version 1.0.0. Values in angle brackets are yours to fill in.

claude mcp add --transport http agent-permissions-index https://agent-permissions-index.vercel.app/mcp

Run in a terminal.

Claude Code permissions

Allow the reads, ask before a change, deny the deletes. Paste into .claude/settings.json.

.claude/settings.json
{
  "permissions": {
    "allow": [
      "mcp__agent-permissions-index__get_index_stats",
      "mcp__agent-permissions-index__get_leaderboard"
    ],
    "ask": [
      "mcp__agent-permissions-index__grade_config"
    ],
    "deny": []
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. It assumes the server is installed as agent-permissions-index. Review it before you use it.