加载中…
LLM evals as MCP tools: score outputs for faithfulness, relevancy, and hallucination.
retriEVAL 是 ChatHome Skills 收录的 MCP 条目，来源为 Official MCP Registry，质量分 73/100。 作为 Model Context Protocol 服务器，它向 Claude、Cursor、Copilot 等兼容 MCP 的 LLM 客户端暴露可直接调用的工具，让 agent 的能力超出模型内置范围。 该条目由 ChatHome Skills 的每日抓取从官方 registry 和 GitHub 发现查询中收录，每次运行都会重新核验仓库事实、新鲜度信号和安装元数据。 质量分综合了官方 registry 收录情况、可用安装命令、许可证完整性、仓库活跃度和 GitHub star 趋势，完整算法说明见来源页。安装命令已列于上方，可直接复制使用。 在将 retriEVAL 加入生产 agent 工作流前，请核对上方的最近更新日期和仓库状态——下方事实快照是最近一次抓取的时点值。
docker run -i --rm ghcr.io/hcarrillo001/retrieval-mcp:1.0.0运行此命令前，请先核对上游源码。
{
"_meta": {
"io.modelcontextprotocol.registry/official": {
"status": "active",
"isLatest": true,
"updatedAt": "2026-08-24T01:28:29.412844Z",
"publishedAt": "2026-08-24T01:28:29.412844Z",
"statusChangedAt": "2026-08-24T01:28:29.412844Z"
}
},
"server": {
"name": "io.github.hcarrillo001/retrieval-mcp",
"title": "retriEVAL",
"$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
"version": "1.0.0",
"packages": [
{
"transport": {
"type": "stdio"
},
"identifier": "ghcr.io/hcarrillo001/retrieval-mcp:1.0.0",
"registryType": "oci",
"environmentVariables": [
{
"name": "ANTHROPIC_API_KEY",
"isSecret": true,
"description": "Judge model API key. Not needed if you set RETRIEVAL_JUDGE_BACKEND=ollama to run the judge locally."
},
{
"name": "RETRIEVAL_JUDGE_BACKEND",
"default": "anthropic",
"description": "Which judge to score with: anthropic (default), openai, or ollama for a fully local judge."
},
{
"name": "RETRIEVAL_JUDGE_MODEL",
"description": "Judge model id, e.g. claude-sonnet-4-6 or qwen2.5:14b for Ollama."
},
{
"name": "RETRIEVAL_BUDGET_USD",
"description": "Spend cap for judge calls. Unset means no cap."
}
]
}
],
"repository": {
"url": "https://github.com/hcarrillo001/retrieval-mcp",
"source": "github"
},
"websiteUrl": "https://retrieval-mcp.com",
"description": "LLM evals as MCP tools: score outputs for faithfulness, relevancy, and hallucination."
}
}