Skip to content

Bench Agent Discovery

by VirajMishra1 · remote · no sign-in to list tools · v0.1.0 · published by GitHub user VirajMishra1

Discover public AI agents, reusable recipes, and trusted benchmark evidence by task.

Read only
CompareRegistry entry

3 tools: 3 only read.

Scanned Oct 21, 2026

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

claude mcp add --transport http bench https://bench.virajmishratakehome.workers.dev/mcp

Tools

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

3 tools

Only reads 3Claude Code: allow
  • get_agentGet one public agent's recipe, public capability manifest, coarse invocation status, owner telemetry, and verified benchmark submissions.No findings

    Why it is in this group

    • The name "get_agent" says "get": a word for reading.
    Description, as the server wrote it
    Get one public agent's recipe, public capability manifest, coarse invocation status, owner telemetry, and verified benchmark submissions.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • list_benchmarksList public, versioned benchmark contracts and only their trusted-runner-verified submissions.No findings

    Why it is in this group

    • The name "list_benchmarks" says "list": a word for reading.
    Description, as the server wrote it
    List public, versioned benchmark contracts and only their trusted-runner-verified submissions.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • search_agentsFind listed public agents by task, capability, category, framework, model, verified evidence, or reuse configuration.No findings

    Why it is in this group

    • The name "search_agents" says "search": a word for reading.
    Description, as the server wrote it
    Find listed public agents by task, capability, category, framework, model, verified evidence, or reuse configuration. Owner telemetry and controlled benchmark evidence are returned separately.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false

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 0.1.0. Values in angle brackets are yours to fill in.

claude mcp add --transport http bench https://bench.virajmishratakehome.workers.dev/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__bench__search_agents",
      "mcp__bench__get_agent",
      "mcp__bench__list_benchmarks"
    ],
    "ask": [],
    "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 bench. Review it before you use it.