FetchGate

Name: FetchGate
Author: MoAz06

@MoAz06

代码仓库文档

关于 FetchGate

The agent only answers from a page it actually read.

#ai-agents #guardrails #retrieval #provenance #fetch #mcp #web-scraping

基本信息

分类

其他

许可证

MIT

运行时

python

发布者

MoAz06

提交者

Mohamed Azahrioui

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "fetchgate": {
      "command": "fetchgate-mcp"
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is FetchGate?

When an AI agent fetches a page and it quietly comes back empty (a JavaScript app, a 403, a Cloudflare wall), the agent often does not stop. It answers from memory and never tells you it did not actually read the page. FetchGate refuses to let that happen: the agent gets the content only when the page was really read, and an honest refusal otherwise.

How it decides

Every fetch gets one verdict: RETRIEVED (a real read), FAILED (a definite non-read), or UNKNOWN (fetched, but not enough readable content, often a JavaScript page). Three layers feed it (transport, extraction, content-validity), and the strictest wins. Thin or empty pages can escalate to a real headless browser and get re-checked.

It is deterministic, needs no API keys, and calls no model.