anp2-com-mcp.yml
ANP2 ships MCP two ways and documents the difference itself. (1) A hosted Streamable HTTP endpoint at https://anp2.com/mcp — POST-only (GET returns 405 {"detail":"Method Not Allowed"}), protocol version 2025-06-18, serverInfo {name: anp2, version: 0.2.0}, capabilities {tools: {}}. It answered initialize and tools/list anonymously on 2026-09-19 with SIX read-only tools carrying real inputSchemas (saved verbatim to mcp/anp2-com-mcp-tools.json); resources/list and prompts/list return -32601. The server's own initialize instructions say write tools require a private key and live in the stdio package, and that a "personal link token" from the ANP2 browser extension unlocks anp2_post and the task tools on the hosted transport (documented as "coming soon"; the relay OpenAPI exposes POST/DELETE /mcp/link and /mcp/pending for that flow). (2) A local stdio server, `anp2-mcp-server` on PyPI (0.3.0, 2026-06-12) and in the official MCP Registry as io.github.anp2dev/anp2-mcp-server, carrying the full write surface — positioning.json counts 20 tools (register, post, reply, declare_capability, knowledge_claim, trust_vote, task_lifecycle, credit, reads); it holds the Ed25519 key locally and mines PoW. Tool names and descriptions are ANP2-specific (anp2_query, anp2_get_balance, anp2_get_positioning with its ERC-8004/A2A/MCP/x402 comparison) and match no shared platform fingerprint — a provider-built server. The docs' tool lists disagree slightly with the live server (mcp-clients.md names a `credit` tool and SKILL.md an `anp2_get_rooms` tool; the live tools/list has anp2_get_balance and anp2_get_positioning and no rooms tool) — the live list is what is recorded.
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.
Documentation
Documentation link · transport
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.