Provides access to the Nvidia Supplier Map dataset, including tools to search supplier edges by tier, supplier, ticker, and honesty flag, and to retrieve dataset version and metadata.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@vhsgreed-mcp show all HBM tier suppliers flagged UNVERIFIED"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
MCP (Model Context Protocol) server exposing vhsgreed's data products to AI
agents. First dataset: Nvidia Supplier Map 2026 — 29 supplier edges across
9 tiers (HBM, foundry, power, ABF substrate, …), each carrying an honesty
flag: disclosed / inferred / estimate / UNVERIFIED.
Free tier returns structure. Paid tier (access token) adds verbatim evidence quotes, source URLs, and filing dates.
dataset_info — version, edge count, tiers, honesty flags, refresh cadence
search_supplier_edges — filter by tier, supplier, ticker, flag.
Optional full_access_token upgrades the response to the full tier.
Related MCP server: LiveDataLink
npm install
npm run build
node dist/index.js{
"mcpServers": {
"vhsgreed": {
"command": "node",
"args": ["/path/to/vhsgreed-mcp/dist/index.js"]
}
}
}Paid tier: set VHSGREED_TOKEN_ENDPOINT (payment rail /validate URL) and
VHSGREED_FULL_TOKEN in the server env to your access token. Purchase:
https://vhsgreed.win/ai/products/ ($10 supplier-edges-full).
Tokens bought via the $5 "denmark-cvr agent access" payment link
(https://buy.stripe.com/eVq8wR51G2Hs6NsayHew806) carry
product: "denmark-cvr" from the rail's /validate. This server recognizes
them as valid tokens (no invalid_token error) but the Nvidia Supplier Map
full tier requires a supplier-edges-full token; a denmark-cvr token gets a
clear wrong_product_token message plus the free tier. A denmark-cvr token
instead grants agent access to the Apify actor
vhsgreed/denmark-cvr-api-fresh: run it directly on Apify or call it from
your MCP client via the vhsgreed server docs in the buyer email.
data/supplier-edges.csv is copied from the deterministic build in
prod/repos/nvidia-supplier-map/ (same pipeline as the Gumroad product and
the /ai/products/ JSON API). Refresh cadence: quarterly; next scheduled
disclosure is NVDA's 10-Q (~late Nov 2026). Bump version in src/index.ts
when the dataset updates.
Free-tier use requires attribution: "vhsgreed.win". Full tier: single-entity use, redistribution prohibited.
Flags are honest. UNVERIFIED means we could not verify the edge; the flag
gradient exists precisely so agents (and humans) can weigh confidence
themselves. We do not blur the difference.
MIT. Data itself: free tier CC-BY (attribution), full tier single-entity.
Karl Sundström · vhsgreed.win · Enskild firma, Org.nr 990607-1619
This server cannot be deployed
US government data as clean JSON for AI agents: SAM.gov contract opportunities, USAspending awards, Grants.gov grants, House STOCK Act trades, and SEC EDGAR filings (Form 4 insider trades, 8-K events, 13F holdings, 13D/G stakes, XBRL fundamentals, 10-K/10-Q sections). 19 read-only tools. Data is as fresh as each source publishes; congressional trades lag up to 45 days and report dollar ranges (House only). Free tier, no card.
Hosted on-chain data for AI agents: DEX trades, OHLCV, top traders, fund tracing, address labels.
SEC filings and financial data for AI agents: 59 tools for statements, valuation and supply chains.
Real SEC, 13F, insider, congress & macro data your AI agent can cite. Hosted MCP, 24 tools.
