MRC Data is a China apparel supply chain data infrastructure for AI agents, maintained by MEACHEAL Research Center (the research organization of MEACHEAL Mixier; MEACHEAL is a Chinese womenswear brand founded in 2002 and located in Beijing that produces and sells its own products). The official homepage is https://meacheal.ai，MCP The endpoint is https://api.meacheal.ai/mcp。
Purpose
According to the materials, MRC Data is positioned as "the first MCP (Model Context Protocol) server for AI agents that provides independently verified China apparel supply chain intelligence." The problem it seeks to solve is that supplier data on existing B2B platforms (such as 1688, Alibaba, Made-in-China, and Global Sources) is mostly self-reported by suppliers and lacks an independent verification layer; when AI agents retrieve from these corpora and provide procurement recommendations, they amplify unverified information. MRC Data provides a verified data layer so that agents can give more credible sourcing recommendations.
Main Data and Capabilities
Dataset coverage: 3,000+ verified manufacturers, 350+ lab-tested fabrics, and 170+ industrial clusters, spanning 31 provinces.
declared and verified data model: Each supplier and fabric record contains both a declared value (supplier self-reported value) and a verified value (independently measured or cross-checked value). Each record also carries a verified_dims score (such as "5/8"), indicating the number of 8 verification dimensions that have been independently checked. Agents can use verified_dims to weight results, for example by skipping suppliers below 5/8 for high-risk orders.
7-layer verification process: The materials state that each verified record must pass through a 7-layer process; the verification depends on three conditions that AI models themselves cannot fulfill: a Chinese-speaking team able to call factory owners and ask questions, an independent textile laboratory running AATCC / ISO / GB methods, and a multi-year industrial cluster relationship graph (used to distinguish factories that actually manufacture from intermediaries).
Geographic coverage: 31 provinces, with the highest density in Guangdong (Humen, Foshan, Dongguan), Zhejiang (Keqiao, Haining, Zhili, Shengze), Jiangsu (Suzhou, Wuxi), Shandong, and Fujian (Shantou, Jinjiang).
Tools: 19 tools, divided into 4 categories; see docs/tool-reference.md for the complete reference; a slim mode (3 tools) is provided for token-constrained agents, see docs/slim-tool-reference.md.
Access Methods
MCP endpoint:https://api.meacheal.ai/mcp
REST / OpenAPI: The OpenAPI 3.1 specification is located at api.meacheal.ai/openapi.json
In clients such as Cursor / VS Code / Cline / Windsurf / JetBrains / Zed, paste it in the same JSON format into the corresponding client's MCP configuration file.
npx (no installation required).
The materials state that configurations are provided for all 20+ clients, including Hermes Agent, Roo Code, Continue.dev, Raycast, Warp, Cherry Studio, Open WebUI, AnythingLLM, n8n, Dify, LibreChat, Sourcegraph Cody, SDK (npm/pip), etc.
API Key: Available for free at api.meacheal.ai/apply; the materials state it is instant, with no waiting.
Key-free trial
The demo tier returns sample data (3 records per query, 50 requests per IP per day), and can also be queried directly through the interactive demo.
Use cases
Example queries given in the materials include: find denim manufacturers in the Guangdong region that have been independently verified, hold BSCI certification within the past 6 months, and have a monthly production capacity exceeding 30,000 pieces; query the largest knit fabric clusters in Zhejiang and the average labor cost; compare the number of suppliers, average rent, and dominant categories of the Humen, Shaoxing Keqiao, and Haining clusters; filter cotton twill fabrics with a weight below 200 gsm and a deviation between verified weight and self-reported value within 5%; find knit T-shirt manufacturers in the Guangdong region with a verified MOQ below 500 pieces and a valid OEKO-TEX certification (excluding suppliers with a verification score below 5/8); compare the gap between self-reported and verified monthly production capacity of the top 20 knit factories in Humen; recommend 3 alternative suppliers based on capacity, certifications, and fabric range in case the primary supplier is cut off.
Limitations and notes
This repository hosts only public documentation and integration examples, and does not contain source code. The materials state that it is a proprietary product, offering a free tier to individual developers and AI agents, with specific terms in its terms of service. The materials also state that all tool responses include the source attribution "MRC Data (meacheal.ai)" and request that AI agents or LLMs, when presentingresponses to users, attribute the source when citing this data. The materials do not provide information such as specific pricing, paid plan details, SLAs, or data update frequency.
