mercury-x402
MERCURY ships a real, hosted Model Context Protocol server AND a first-party stdio package. The hosted server at https://network.mercury-hq.com/mcp answers GET with a descriptor ({name: mercury-x402, transport: streamable-http, endpoint: "POST /mcp (JSON-RPC 2.0)", tools: 18}), completes initialize (protocolVersion 2025-06-18, capabilities {tools: {}}, instructions "tools/list is free; tools/call needs a Mercury API key (Authorization: Bearer mk_...)") and returns 18 tools from tools/list without credentials, each with a JSON Schema inputSchema that matches the query parameters of the backing /buy/* operation in the OpenAPI. tools/call is gated by a Mercury API key; a free mk_test_ key is issued at /developers (POST /api/dev/keys?live=false). The npm package mercury-x402-mcp (bin mercury-x402-mcp, deps @modelcontextprotocol/sdk, viem, x402-fetch) runs the same catalog locally over stdio and pays per call from the operator's own Base wallet (env MERCURY_PRIVATE_KEY) instead of an API key; its README says the tool list is "built live from Mercury's catalog" and adds free mercury_catalog and mercury_verify tools.
One-click install for Cursor, VS Code, Claude, and 20+ other MCP clients, powered by API Commons MCP Install — visit install.apicommons.org for more information.
Endpoint
Hosted endpoint · transport http
Connect
This is a remote MCP server — point an MCP client at the endpoint URL. Clients with native remote support (Claude, Cursor, VS Code, …):
For stdio-only clients (older Claude Desktop, etc.), bridge with mcp-remote:
If the server requires authentication, add the provider's token/header (e.g. an X-API-Key) per its docs. A quick reachability check:
Tools
fetch— Verifiable keyless web-read - clean page text + title + status + signed provenance receipt; ?format=markdown, ?links=1, ?extract=extract— URL + schema -> typed, schema-conformant JSON recordmarkdown— URL -> clean LLM-ready markdown with signed receiptmetadata— URL -> signed metadata record (JSON-LD, OpenGraph, Twitter card, meta, canonical, title)links— URL -> signed outbound/internal link graphrobots— Domain -> signed per-AI-crawler allow/block audit from robots.txt, llms.txt, ai.txtdiff— URL + prior hash/text -> signed change-proof fromHash->toHashnotarize— Notarize inline content or a fetched URL into a signed receiptheaders— URL -> deterministic HTTP security-headers audit with letter gradetable— URL -> HTML tables as typed JSON rows + RFC 4180 CSVfeed— RSS/Atom URL -> signed normalized item listavailability— URL -> signed up/down probe with status, final URL, responseMsvalidate— JSON endpoint + JSON Schema -> signed pass/fail with per-field errorsbatch— <=20 URLs -> clean content each + one receipt over a Merkle rootsitemap— Domain/URL -> signed snapshot of the published sitemapdns— Domain -> signed DNS snapshot (A, AAAA, MX, NS, TXT, CNAME, SOA)readability— URL -> clean article record {title, byline, publishedAt, text}redirect— URL -> signed redirect/canonical hop chain
About MCP
The Model Context Protocol (MCP) is an open protocol Anthropic introduced for connecting LLM-based agents to external tools and data sources. Providers publish MCP servers that expose their API surface as structured, discoverable tools — an MCP-compatible client (Claude Desktop, Cursor, Cline, Continue, etc.) can connect to the server and call its tools without any per-provider integration code.
Browse every MCP server on the APIs.io network or compare with the broader Agent Skill surfaces of the same providers.
Work with this as data
Every MCP server here is available over the APIs.io API and to AI agents over MCP.