sverklo-bench

sverklo/sverklo-bench
★ 2 stars TypeScript 🤖 AI/LLM Updated today
bench:primitives — the public, reproducible benchmark for code-intelligence MCP servers. 90 tasks across 3 datasets, 5 baselines. Companion to github.com/sverklo/sverklo.
View on GitHub →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "sverklo-bench": {
      "command": "npx",
      "args": [
        "-y",
        "sverklo/sverklo-bench"
      ]
    }
  }
}

README Excerpt

A public, reproducible benchmark for code-intelligence MCP servers and code-search baselines. **90 hand-verified tasks** across **3 OSS codebases** (sverklo, express 4.21.1, lodash 4.17.21), **4 task categories** (definition lookup, reference finding, file dependencies, dead code), **5 baselines** (naive grep, smart grep, [sverklo](https://github.com/sverklo/sverklo), [jcodemunch-mcp](https://github.com/jgravelle/jcodemunch-mcp), [GitNexus](https://github.com/abhigyanpatwari/GitNexus)).

Topics

ai-codingbenchmarkclaude-codecode-intelligencecode-searchcursormcp