datapulse_dns_searchlabels

Description

Search the DataPulse database for domain labels similar to a search term.

This is the primary tool for finding typosquatting, brand impersonation, and lookalike domains. Returns the PSL+1 label — the registrable label one level below a public suffix (google in google.com, example in example.co.uk) — and the TLDs where that label is listed in the corpus, giving a complete picture of a string’s presence across the domain namespace. Hostnames and subdomains are not indexed and cannot be searched.

Matches domain-name LABELS only — does NOT search WHOIS/RDAP record contents. It cannot find domains by registrant email/name/organization (no reverse-WHOIS; see datapulse_help(topic="registrant_lookup")).

Search Behavior:

Fuzzy matching: built for lookalikes — homoglyphs and IDN/confusable characters. It does NOT systematically find plain typos (omitted, transposed or adjacent-key letters): for paypal, the fuzzy list filled with IDN variants and left out paypl, which an exact search finds. To check typo variants, generate them yourself and search each with filter=[“exact”]
Match types: ’exact’, ‘substring’, ‘fuzzy’. Omit for all types
Score is Levenshtein edit distance (0 = exact, 1 = one edit away, etc.), even for IDN/homoglyph matches
Results ordered by increasing score — nearest matches first

Use Cases:

Find typosquatting domains
Discover brand variants and impersonation
IDN homograph abuse detection
Defensive registration audit

Response: CSV with header row then label,type,score,tlds data rows, at most 100 labels. Up to two lines can precede the header: error: ... (the API’s own error text) and complete: false (...) when the label limit was reached or the API cut its answer short — more labels then exist. The tlds column is the pipe-delimited, COMPLETE list of every TLD where the label is listed (no cap). The corpus lists active, functioning names only, so a TLD missing from the list does not mean the name is unregistered there — it may be registered and parked, unresolving or new. Treat the list as the full set of listed TLDs: do NOT add TLDs, domains, ownership, or market/defensive annotations that are not present in the tool output — registration in a TLD does not establish who owns it (see datapulse_help(topic="registrant_lookup")). Very common strings may produce result sets large enough to time out; use filter to narrow the match types if this happens.

For usage guidance, call datapulse_help(topic="searchlabels")

Parameters

Parameter	Type	Description
`search_term` required	`string`	Search term to query. Accepts any valid UTF-8 string. Min length: `3`Max length: `255`
`filter`	`string[]`	Restrict to listed match types.

Input schema (JSON)

{
  "additionalProperties": false,
  "properties": {
    "filter": {
      "description": "Restrict to listed match types.",
      "items": {
        "enum": [
          "exact",
          "substring",
          "fuzzy"
        ],
        "type": "string"
      },
      "type": "array"
    },
    "search_term": {
      "description": "Search term to query. Accepts any valid UTF-8 string.",
      "maxLength": 255,
      "minLength": 3,
      "type": "string"
    }
  },
  "required": [
    "search_term"
  ],
  "type": "object"
}

Generated from the live server (DataPulse MCP 1.0.0) on October 7, 2026.