My Lead Fox MCP server

Showing 6 of 6 tools

• myleadfox_column_values

  Read-only

  The values a dropdown or tags column holds, spelled exactly as My Lead Fox stores them -- "Facebook", not "facebook". Use it before filtering on a tags or dropdown column: values are case-sensitive, and a value that does not exist does not raise an error, it quietly matches almost nothing. Returns a ready-made example filter using a real value. For the yes/no columns the list is complete. For every other column it is a SAMPLE of values seen in real data, because My Lead Fox has no endpoint that lists distinct values -- an unlisted value may still exist. Requires col. FREE: makes no API call and spends no credits. Costs no credits and makes no API call.

• myleadfox_enrich_domains

  Read-only

  Find email and phone contacts for domains already in the My Lead Fox dataset. THE SCARCEST CREDIT POOL BY A LONG WAY -- a typical account holds a few hundred enrich credits against hundreds of thousands of download credits. Do not loop this over a large list. Enrichment is synchronous and each domain is checked live, so keep batches small; roughly one 100-domain batch per minute is the sustainable rate. Every domain comes back in exactly one of SIX outcomes and only `enriched` is charged. Read them before concluding anything: `no_contact_found` means the shop genuinely has no contacts, but `skipped_no_credit` means the balance ran out and `not_attempted` means the call timed out -- neither is evidence about the shop, and `not_attempted` domains should simply be sent again. COSTS: spends the "enrich" credit pool.

• myleadfox_get_rows

  Read-only

  The only tool that returns shop data. Page a searchId from myleadfox_search, or pass filters inline. Any page is directly addressable -- page 500 needs no walk through 1-499. Filters -- one per condition, joined by `conn`. Call myleadfox_list_columns first: each column takes exactly one kind, and using the wrong one is refused. The operator is always `op`, one comparand is always `value`, a list is always `values`. text: {"kind":"text","col":"name","op":"contains","value":"coffee"} · numeric: {"kind":"numeric","col":"products","op":"gte","value":100} or op "between" with min and max · dropdown: {"kind":"dropdown","col":"country","values":["US","GB"]} · tags: {"kind":"tags","col":"pixels","op":"none","values":["Facebook"]} · date: {"kind":"date","col":"registered","from":"2024-01-01"}. Dropdown and tags values are CASE-SENSITIVE and a wrong-case value does not fail -- "facebook" matches 3 shops where "Facebook" matches a million. Check spellings with myleadfox_column_values. CHARGED PER ROW DELIVERED, up to 100 per call. Ask for `columns` you actually need. If the download balance is smaller than the page, you receive as many rows as you can afford and the result says `truncatedByCredits: true` -- that means the balance ran out, NOT that only that many shops matched. Never report a truncated page as the full result. A null value means the field has never been measured for that shop. It does not mean zero and does not mean empty. COSTS: spends the "download" credit pool.

• myleadfox_list_columns

  Read-only

  List every queryable column with the ONE filter kind it takes, and how each kind is written. Call this before building filters: each column takes exactly one kind -- `country` is dropdown, not text; `pixels` is tags; `has_email` is dropdown with values yes/no -- and using the wrong kind is refused. Dropdown and tags columns include a preview of their values; call myleadfox_column_values for the full list. FREE: makes no API call and spends no credits. My Lead Fox has no discovery endpoint, so this list and its kinds come from the vendor's documentation, checked against a live response. Costs no credits and makes no API call.

• myleadfox_lookup_domains

  Read-only

  Fetch shop records for up to 100 domains you already have. Use this instead of filtering when you know the domains. CHARGED PER DOMAIN FOUND -- misses cost nothing and are named. THREE lists come back and they mean different things. `rows` were found. `missing` are not in the My Lead Fox dataset and will still be missing tomorrow. `withheldForCredits` ARE in the dataset but past your download balance -- they are real shops you simply could not afford this call, and reporting them as not found would be wrong. Top up and ask for those again. A null field means never measured, not zero. COSTS: spends the "download" credit pool.

• myleadfox_search

  Read-only

  Count how many shops match a filter set, WITHOUT returning any rows. Returns a total and a searchId you can then page with myleadfox_get_rows. Filters -- one per condition, joined by `conn`. Call myleadfox_list_columns first: each column takes exactly one kind, and using the wrong one is refused. The operator is always `op`, one comparand is always `value`, a list is always `values`. text: {"kind":"text","col":"name","op":"contains","value":"coffee"} · numeric: {"kind":"numeric","col":"products","op":"gte","value":100} or op "between" with min and max · dropdown: {"kind":"dropdown","col":"country","values":["US","GB"]} · tags: {"kind":"tags","col":"pixels","op":"none","values":["Facebook"]} · date: {"kind":"date","col":"registered","from":"2024-01-01"}. Dropdown and tags values are CASE-SENSITIVE and a wrong-case value does not fail -- "facebook" matches 3 shops where "Facebook" matches a million. Check spellings with myleadfox_column_values. Do this first. Rows are charged per row delivered, so sizing a segment before downloading it is how you avoid spending download credits discovering that a filter matched two million shops. Charges 1 search credit even when nothing matches. Use myleadfox_list_columns to see what can be filtered on. COSTS: spends the "search" credit pool.