SchemaFetch MCP Server
Enables AI agents to extract structured JSON data from web pages using presets or custom JSON schemas, with confidence scores and source snippets, paid per call via USDC on Base.
README
SchemaFetch
Schema-constrained web extract for AI agents. Pass a URL + preset (or JSON schema), get structured JSON back with confidence and source snippets. Pay per call in USDC on Base via x402 — no API keys for buyers.
Separate product from ReceiptStamp. Reuses the same Cloudflare account, CDP facilitator keys, and Base payTo wallet; does not merge codebases.
Live (after deploy): https://schemafetch.panmediatech.workers.dev
API
| Route | Auth | Purpose |
|---|---|---|
GET /health |
free | Liveness + preset list |
GET /v1/sample?preset=contact |
free | Fixed demo HTML extract |
GET /v1/presets |
free | Preset schemas |
GET /openapi.json |
free | OpenAPI doc |
POST /x402/extract |
$0.02 USDC | { "url", "preset" } or { "url", "schema" } |
GET /x402/extract?url=&preset= |
$0.02 USDC | Agentic.Market GET compatibility |
POST /x402/extract/custom |
$0.04 USDC | { "url", "schema" } only |
Presets
saas_pricing · contact · team · product · faq
Example (after payment)
POST /x402/extract
{
"url": "https://example.com/pricing",
"preset": "saas_pricing"
}
{
"data": {
"plans": [{ "name": "Pro", "price": "$29 / mo", "period": "month", "features": [] }],
"currency_hints": ["USD"]
},
"confidence": 0.7,
"sources": [{ "field": "plans", "quote": "..." }],
"fetched_at": "2026-07-11T00:00:00.000Z",
"cache": "miss",
"preset": "saas_pricing"
}
Structure
src/extract.js— pure extract logic (heuristics + presets)src/worker.mjs— Hono Worker + x402 paywall (CDP facilitator)mcp/server.mjs— MCP bridge (free tools + paid extract via x402-fetch)skills/schemafetch/SKILL.md— agent skill (when/how to call; free install)test/— extract + worker tests (402 without real CDP settlement)scripts/bootstrap-x402-payment.mjs— one-time Bazaar first-settlement pay
MCP + skill (distribution, same monetization)
Packaging only — agents still pay the Worker per call. No second price.
You (the seller) do not need to keep npm run mcp open. The product is the live Worker + marketplace listing. MCP is how other people plug SchemaFetch into Claude Desktop / Cursor.
For buyers (your audience)
Step-by-step: mcp/FOR_USERS.md
One-liner config (Claude Desktop / similar hosts):
"schemafetch": {
"command": "npx",
"args": ["-y", "github:Fourier18/SchemaFetch", "mcp"],
"env": {
"SCHEMAFETCH_BASE_URL": "https://schemafetch.panmediatech.workers.dev",
"X402_CLIENT_PRIVATE_KEY": "0xYOUR_BASE_PRIVATE_KEY_WITH_USDC"
}
}
Skill
skills/schemafetch/SKILL.md — instructions agents can load (when to call, presets, budget).
Marketplace (agent HTTP discovery — already live)
https://agentic.market/services/schemafetch-panmediatech-workers-dev
Local
npm install
npm test
Deploy
Same Cloudflare account as ReceiptStamp (panmediatech workers.dev).
# Secrets — file redirect only on this machine (interactive paste can corrupt)
npx wrangler secret put CDP_API_KEY_ID < secrets/cdp_id.txt
npx wrangler secret put CDP_API_KEY_SECRET < secrets/cdp_secret.txt
npx wrangler deploy
Then once, with a funded Base USDC wallet, bootstrap Bazaar indexing:
X402_CLIENT_PRIVATE_KEY=0x... node scripts/bootstrap-x402-payment.mjs
Kill criterion
Zero non-test paid calls by 2026-07-25 = stop funding this product. Listing and self-payments do not count.
Relationship to ReceiptStamp
| ReceiptStamp | SchemaFetch | |
|---|---|---|
| Job | Notarize payload | Extract structured page data |
| Rails | ACP + x402 | x402 only (v1) |
| Worker | receiptstamp |
schemafetch |
| payTo | 0x7e05…dc20 |
same (by design) |
| CDP keys | Worker secrets | same account, new Worker secrets |
See PLAN.md for status and deploy notes.
Built with
| Language | JavaScript (Node.js ≥20, ESM Worker) |
| Runtime | Cloudflare Workers |
| HTTP | Hono |
| Payments | x402 + Coinbase CDP facilitator (USDC on Base) |
| Agent tools | MCP (@modelcontextprotocol/sdk), x402-fetch |
| AI (authoring) | Grok 4.5 via Grok Build CLI v0.2.93 (xAI SuperGrok / Grok Build agent stack) |
Human owns product decisions, wallets, and accounts. Stack details also in TOOLS.md.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.