中国 AI 合规与备案 MCP:生成式AI备案、AI生成内容标识、深度合成、算法备案的条文级问答与自查清单(中文)
- ✓Open-source license (MIT)
- ✓Actively maintained (<30d)
- ✓Clear description
- ✓Documented (README)
git clone https://github.com/savantcat/savantcat-ai-compliance-mcpTools overview
# 中国 AI 合规与备案 MCP · savantcat-ai-compliance
把中国 AI 法规做成 **Agent 可直接调用** 的 MCP 服务:**每条结论都标注法规条文依据**。
- 远程端点(免 Key):`https://savantcat.cn/mcp-compliance`
- 条文级网页版:`https://savantcat.cn/answers/compliance/`
- Smithery:`https://smithery.ai/server/@savant0196/savantcat-ai-compliance`
覆盖 4 部法规 + 1 个强制性国标,共 **42 条**条文级原子问答:
| 集群 | 内容 | 条数 |
|---|---|---|
| 生成式AI暂行办法 | 训练数据、内容安全、个人信息、投诉处置、安全评估与算法备案 | 12 |
| AI生成合成内容标识 | 显式/隐式标识、平台核验、不得删除篡改 | 10 |
| 深度合成 | 换脸/AI配音/数字人的实名、标识、生物识别单独同意 | 8 |
| 算法推荐与备案 | 备案时限与公示、用户关闭选项、反大数据杀熟、安全评估 | 8 |
| 备案实操 | 哪些情形触发备案与评估、需要交什么材料 | 4 |
法规原文来源:中国政府网、中央网信办(cac.gov.cn)公开发布版本;《生成式人工智能服务管理暂行办法》《人工智能生成合成内容标识办法》《互联网信息服务深度合成管理规定》《互联网信息服务算法推荐管理规定》,以及强制性国标 GB 45438-2025(标识方法)。
## 工具(6 个,全部只读)
| 工具 | 作用 |
|---|---|
| `list_topics` | 列出全部条目(可按集群过滤) |
| `search_compliance` | 关键词检索,返回结论 + 条文依据 |
| `get_requirement` | 取单条完整内容(正文 + 依据 + 落地动作 + 常见追问) |
| `self_check` | 按场景汇总成自查待办清单 |
| `filing_route` | 输入服务形态,输出应办事项 + 条文依据 |
| `regulation_info` | 收录法规清单(发布/施行日期、条数、官方原文链接) |
## 用它回答的问题
- 上线 AI 客服 / 智能问答,要不要备案?要不要做安全评估?
- 用别人家的大模型 API 做产品,备案义务在我还是在模型方?
- AI 生成的文案、图片、视频要不要打标?显式标识和隐式标识分别是什么?
- 换脸、AI 配音、数字人做营销素材,要注意什么?
- 算法备案和生成式 AI 备案是一回事吗?多久要办完?
## 本地运行
```bash
pip install -r requirements.txt
python server.py # stdio(桌面客户端)
python server.py --selftest # 不走协议,直接打全部工具
python server.py --transport http --host 127.0.0.1 --port 8766 --stateless
```
## 接入示例
```json
{
"mcpServers": {
"cn-ai-compliance": {
"type": "streamable-http",
"url": "https://savantcat.cn/mcp-compliance"
}
}
}
```
## 数据来源与边界
- 语料只来自**官方公开发布**的法规原文,逐条标注条文号,引文与原文逐字一致(`build_data.py` 内置逐字接地校验,条号或引文对不上直接构建失败)。
- 回答用于**企业自查参考**,不构成法律意见;申报口径以属地网信部门要求为准。
- 本仓库**不含任何客户语料**。
## License
Apache-2.0
What people ask about savantcat-ai-compliance-mcp
What is savantcat/savantcat-ai-compliance-mcp?
+
savantcat/savantcat-ai-compliance-mcp is tools for the Claude AI ecosystem. 中国 AI 合规与备案 MCP:生成式AI备案、AI生成内容标识、深度合成、算法备案的条文级问答与自查清单(中文) It has 0 GitHub stars and its last recorded update is dated 2026-09-18.
How do I install savantcat-ai-compliance-mcp?
+
You can install savantcat-ai-compliance-mcp by cloning the repository (https://github.com/savantcat/savantcat-ai-compliance-mcp) or following the README instructions on GitHub. ClaudeWave also provides quick install blocks on this page.
Is savantcat/savantcat-ai-compliance-mcp safe to use?
+
Our security agent has analyzed savantcat/savantcat-ai-compliance-mcp and assigned a Trust Score of 87/100 (tier: Trusted). See the full breakdown of passed checks and flags on this page.
Who maintains savantcat/savantcat-ai-compliance-mcp?
+
savantcat/savantcat-ai-compliance-mcp is maintained by savantcat. The last recorded GitHub activity is dated 2026-09-18, with 0 open issues.
Are there alternatives to savantcat-ai-compliance-mcp?
+
Yes. On ClaudeWave you can browse similar tools at /categories/tools, sorted by popularity or recent activity.
Deploy savantcat-ai-compliance-mcp to your cloud
Ship this repo to production in minutes. Each platform spins up its own environment with editable env vars.
Maintain this repo? Add a badge to your README
Drop the badge into your GitHub README to show it's tracked on ClaudeWave. Each badge links back to this page and reflects the live Trust Score.
[](https://claudewave.com/repo/savantcat-savantcat-ai-compliance-mcp)<a href="https://claudewave.com/repo/savantcat-savantcat-ai-compliance-mcp"><img src="https://claudewave.com/api/badge/savantcat-savantcat-ai-compliance-mcp" alt="Featured on ClaudeWave: savantcat/savantcat-ai-compliance-mcp" width="320" height="64" /></a>More Tools
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.
An AI skill that provides design intelligence for building professional UI/UX across multiple platforms.
🪨 why use many token when few token do trick. Viral skill + proxy for coding agents that cuts 65% of tokens by talking like a caveman.
CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies
The fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (or native) format with cost tracking, guardrails, load balancing, and logging [Bedrock, Azure, OpenAI, Anthropic, OpenAI, VertexAI, vLLM, Nvidia NIM]
Use Claude Code, Codex, Pi, and OpenCode (and 6 other harnesses) for free (1.3B+ free tokens) from your terminal, app, IDE, or phone, and now from the browser with native browser sessions (multi-harness + multi-model) like OpenClaw (voice supported + ToS friendly)