numbers-online-mcp.yml
Numbers Online operates ONE hosted, stateless, read-only MCP server at POST https://numbers.online/api/v1/mcp (JSON-RPC 2.0 over Streamable HTTP; the provider and Vapi call the transport literal "shttp"). Anonymous initialize and tools/list succeed (200); tools/call is fail-closed behind an API key with the mcp use case. GET returns 405 (response-only, no server-initiated SSE). Nothing to install: the provider ships no stdio package.
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 streamable-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:
Tool quality observed · not scored
tools with readOnly / destructive hints
parameters described (7 total)
state-changing tools without destructiveHint
tool-list tokens (4 tools)
tools over ~1,200 tokens
From an anonymous initialize + tools/list capture measured 2026-10-20; tokens are chars/3.6 of each tool definition. Recorded for the Kin Score's agent readiness as an observed signal — it contributes no points yet.
Tools
phone_lookup— Return a supplementary, low-confidence intelligence bundle for a phone number: validity, formatting, line type, range carrier, country, caller name (CNAM, where available), STIR/SHAKEN verstat (supply the call's verstat to fold it into the spam signal), a labeled spam signal, and a first-party do-not-contact (DNC) signal — a suppression preference registered by the number's own owner, never a copy of any government registry. Advisory only — the agent keeps every routing and dialing decision.line_type— Return only the deterministic fields for a number — validity, line type (mobile / fixed line / VoIP / …), range carrier, country, and formatting. No paid dip; a free, fast classification a voice agent can call before deciding to enrich.caller_risk— Return a labeled, low-confidence risk read for an inbound caller: a spam signal (0–100, higher = riskier), verstat passthrough, and a first-party do-not-contact (DNC) signal — without a name dip. A supplementary input to the agent's own policy, never a verdict on whether to accept or reject the call. Supply the call's STIR/SHAKEN verstat to fold it into the signal — a failed validation raises the risk.dnc_check— Return a first-party do-not-contact signal (suppress / no_match / unknown) for a number, sourced ONLY from suppression preferences registered and verified by the number's own owner inside Numbers Online — never a copy of any government or licensed do-not-call registry. "suppress" = the owner asked not to be contacted on this channel; "no_match" = no suppression preference on record (absence of suppress is not consent); "unknown" = the number could not be evaluated. Comes with a signed receipt recording the status "as of T". Advisory input to the agent's own process — it does not assert that a call is lawful.
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.