# Fredyee Hallucc MCP server

Source: https://glimind.com/server/mcp-registry%2Fio.github.fredyee%2Fhallucc
Registry id: `mcp-registry/io.github.fredyee/hallucc`
Status: live (checked 2026-10-27T09:03:36.494Z)

## Verdict

Fredyee Hallucc MCP server answered 99.92% of Glimind's liveness checks over 30 days, better than 31% of MCP servers, with 844 ms median handshake and 2 outages in 90 days. Its agent-readiness grade is A (100/100) across 5 tools, costing an agent about 2,030 tokens to load. Checked every few minutes; no tool is ever called.

## Measurements (30 days)

| Metric | Value | Better than |
|---|---|---|
| Uptime, 30 days | 99.92% | 31% of MCP servers |
| Median handshake latency | 844 ms | 24% of MCP servers |
| p95 handshake latency | 3449 ms | — |
| Checks in 30 days | 2,810 | — |
| Outages, 90 days | 2 | — |
| Median recovery | 2 min | — |
| Agent-readiness score | 100/100 (A) | 93% of MCP servers |
| Tools | 5 | — |
| Tool-list size for an agent | ~2,030 tokens | 52% of MCP servers |
| Days observed | 27 | — |

## Install

```json
{
  "mcpServers": {
    "fredyee-hallucc-mcp-server": {
      "type": "http",
      "url": "https://aihcc.cloud/mcp"
    }
  }
}
```

Claude Code:

```sh
claude mcp add --transport http fredyee-hallucc-mcp-server https://aihcc.cloud/mcp
```

## Tools (5)

- `check_cua_actions`: Computer-Use Agent 动作风险分级。提交动作轨迹 → 逐步返回 L0(放行)/L1(放行+记录)/L2(需确认)/L3(阻断)/takeover(凭据场景挂起等用户亲手输入) 裁决、matched_rules、outcome，及 L0-L3 汇总计数；响应含 ruleset_version（规则库版本）
- `check_cua_code_audit`: CUA 代码审计：静态分析 Agent 源码，检测危险导入（subprocess/os.system 等）、权限边界（eval/exec）、注入面（字符串拼接 shell）、危险默认值、沙箱缺失。纯规则扫描，不调 LLM、不耗额度。支持单文件或多文件扫描。
- `check_safety`: AI 安全网关检测。对 prompt + output 跑 OWASP LLM Top 10 三层检测：40+ 特征规则引擎（注入/越狱/有害内容/敏感信息泄露/欺诈）+ 可选 LLM 深度分析 + 可选幻觉检测。返回 passed、risk_score、risk_level、threats[]（category/se
- `verify_agent`: Agent 输出与轨迹自检。对最终回答跑逐声明幻觉核验，并对执行轨迹做六维结构化评估（事实性/来源/指令合规/工具声明一致/任务完成/反思）。返回 main.claims、steps[].checks、quota。耗 detect 额度。
- `verify_text`: 逐声明幻觉核验。输入文本 → 提取声明 → 逐声明验证（supported/refuted/unverified）+ 来源引用。返回红/黄/绿汇总、每条声明的状态/置信度/理由/来源、citations、quota。耗 detect 额度（与 Web 端同源）。

Measured by Glimind (https://glimind.com): MCP handshake + tool list only, no tool is ever called. CC BY 4.0.