Wanhe Compliance — Chinese Content Compliance Checker
Offline MCP server that flags Chinese ad-law banned terms and checks AI content labeling
- Pricing
- Open source
- Platforms
- CLI, self-hosted
- Version
- 1.1.5
- Listing updated
- 1 day ago
Similar MCP servers
About Wanhe Compliance — Chinese Content Compliance Checker
Wanhe Compliance — Chinese Content Compliance Checker is an open-source MCP server for security. It is available on CLI and self-hosted.
The README describes an MCP server that gives AI agents a Chinese content compliance check for marketing copy such as product pages, short-video scripts and public account articles. It targets the phrases that commonly break Chinese advertising rules, like absolute claims, medical or cure claims and guaranteed returns, and reports each hit with the matched text, the cited article and a suggested edit.
According to the README, the server runs locally over stdio and works with Claude Desktop, Cursor and Cline. It falls back to a built-in JSON-RPC implementation when the official mcp library is missing. The project says the free version sends no data off the machine, and its rule engine is deterministic, so the same input returns the same output.
It is aimed at Chinese-language marketing and content teams and their developers who want compliance checks inside an AI writing workflow. The README states that results are rule matches and citations, not legal advice.
Tools this server provides (3)
Taken from the project documentation; the running server may differ.
- check_content
- Checks text for compliance risks and returns each issue with its basis and suggestion.
- check_ai_label
- Determines whether text needs explicit or implicit AI-generated content labelling.
- self_assessment
- Runs a 17-item organisational compliance questionnaire and returns a coverage score.
What it does
- Banned term detection
- Scans Chinese text for advertising-law banned phrases and returns each hit with its location and suggested fix.
- Legal basis citations
- Each finding names the article it relies on, such as Advertising Law Article 9 or 17.
- AI content label check
- Assesses whether text needs explicit or implicit AI-generated content labelling under the cited Chinese measures.
- Organisation self-assessment
- A 17-item compliance questionnaire that returns a coverage score and a remediation list.
- Offline local operation
- The free version runs on the local machine with no network calls and no usage data collection, according to the README.
Good for
- Checking e-commerce product copy for absolute claims before publishing
- Reviewing short-video scripts for medical or efficacy wording
- Deciding whether AI-generated Chinese articles need a visible AI label
- Running an internal compliance questionnaire for a team using generative AI
Pricing
Open source
MIT-licensed free local version; a cloud version with a fuller rule set is mentioned in the README without public pricing.
Prices change. Check wh-bi.github.io before you buy.
Wanhe Compliance — Chinese Content Compliance Checker FAQ
Does the free version send content off the machine?
The README says the free version runs fully locally, content does not leave the machine, and it makes no network requests unless the user configures the cloud version.
Which clients can use it?
The README names Claude Desktop, Cursor and Cline, and says any MCP client using stdio transport can be configured the same way.
Is the output legal advice?
No. The README states that the tool performs rule matching and citations of legal articles and does not constitute legal advice.
Is Wanhe Compliance — Chinese Content Compliance Checker free?
Yes. Wanhe Compliance — Chinese Content Compliance Checker is open source, so you can use and self-host it for free.
What are the best alternatives to Wanhe Compliance — Chinese Content Compliance Checker?
The closest alternatives to Wanhe Compliance — Chinese Content Compliance Checker are Wanhe AI Compliance Checker, AIM SHIELD, Passiv Web Tools, 中国 AI 合规与备案(条文级自查) and Proofof Ai MCP.
Where this comes from
- Description: written from the project's README, the project's website and its MCP Registry entry.