Niches · CRM & customer platforms
Find the right record before acting
One company's docs show a full fit.
What counts as done
The matching record ids across contacts, companies and deals, or an explicit none; current open records are not hidden by default filters; a just-created record is findable.
Measured, for our own tests: the ids come in one or two calls, and a just-created record is findable within 10 seconds. No docs we read state these figures, so they don't count against a listing here.
How agents ask for it: is this person already in the CRM
, find the lead with this phone number
, which deal is this email about
, look up Acme's open opportunities
Does the job 1
The docs show it meets the test above, and nothing we read says otherwise.
Does the job, with a limit 19
It does the job, but something in the docs or a user's account will get in an agent's way. The limit is named in each row.
| Listing | Why |
|---|---|
| Affinity B Project Affinity, Inc. | Separate search tools for persons, companies and opportunities, plus a semantic search endpoint; the search endpoints are beta and no indexing delay after a write is stated. source |
| Attio API + MCP B Attio Limited | search-records does full-text search across people, companies and deals by name, email or domain in one call, but phone normalisation and any indexing delay after a write are not stated. source |
| Breakcold C Logike SAS | records_list takes a search term and inbox search matches names, emails and phone numbers, but whether records search spans every object type, normalises phones, or lags writes is not stated. source |
| Capsule CRM C Zestia Limited | list_contacts filters on email, name and custom fields with is and contains operators, and REST has separate party, opportunity and project searches; there is no single cross-object search and no indexing delay is stated. source |
| Close API + MCP C Elastic Inc | find_contacts matches exact emails and secondary entries, normalises phones to the user's country, and list_objects spans leads, contacts and opportunities, but no indexing delay after a write is stated. source |
| Copper API D Copper CRM, Inc. | Per-entity search endpoints with documented phone normalisation (partial and fuzzy), and fetch person by email, but no single search across people, companies and opportunities and no stated indexing delay. source |
| Microsoft Dynamics 365 Sales B Microsoft Corporation | search_data searches structured and unstructured data across tables and read_query runs SQL SELECT, but search_data is billed per use in Copilot Credits and no indexing delay is stated. source |
| folk API + MCP C Folk Inc. | Separate search tools for people, companies and objects filter on emails and phones with eq and like, but there's no single cross-object search, no phone normalisation is stated, and no indexing delay is given. source |
| Freshsales API E Freshworks Inc. | /api/search spans leads, contacts, accounts and deals in one call and /api/lookup matches an exact email, but results stop at 100 and phone normalisation and indexing delay aren't stated. source |
| HubSpot API + MCP BB HubSpot, Inc. | search_crm_objects filters any object type with up to five filter groups, on the CRM search API with no vector search, and a17 shows search lagging a fresh write by seconds. source |
| Nutshell D Nutshell, Inc. | Separate MCP searches for leads, companies and people, following the web app's visibility rules; no cross-object search, phone normalisation or indexing delay is documented. source |
| Payani AI F Payani Media LLC | Contacts, companies and deals can be looked up, but the tool list sits behind OAuth, so search fields, matching rules and indexing delay can't be read. source |
| Pipedrive API + MCP B Pipedrive Inc. | /itemSearch searches deals, persons, organisations and leads in one call with exact_match, but search has its own 10-requests-per-2-seconds limit, and a24 shows a community MCP hiding current deals behind default filters. source |
| Salesflare E Salesflare BVBA | /contacts filters by email, phone_number and domain and /accounts by domain, but each object is searched separately and phone normalisation and indexing delay aren't stated. source |
| Salesforce API + MCP F Salesforce, Inc. | A find tool sits beside soqlQuery in the standard servers (a36), but a12 shows a voice agent failing to match leads by phone number, and the search docs could not be read. source |
| Streak API + MCP E Rewardly, Inc. | The MCP searches pipelines, boxes, contacts and organisations, but how search matches emails or phones and any indexing delay are not documented, and the tool list isn't published. source |
| SugarAI (SugarCRM) D SugarAI Software Inc. | /globalsearch searches across modules and /:module/filter takes exact filters, but the index is Elasticsearch with an admin-set refresh interval, so a fresh record can lag, and phone normalisation isn't stated. source |
| Twenty API + MCP C Twenty.com PBC | find_records filters any object and direct mode adds per-object find tools, but each object is searched separately and phone normalisation and indexing delay aren't documented. source |
| Zoho CRM BB Zoho Corporation Private Limited | Search takes email and phone across every email and phone field, states its default filters (non-converted, approved only) and returns more_records, but searches one module at a time and documents an indexing delay right after a write. source |
What stops agents finishing
- search hides current records behind undocumented default filters (a24)
- phone or email formats not normalised, so lookups miss (a12)
- search index lags writes (a17)
Around it
Before: voice-agents.tool-call-mid-call, personal-assistants.inbox-triage-draft, support.research-context. After: crm.upsert-contact, crm.log-activity, crm.move-deal.
Where this comes from
Weight 11 of 100 in CRM & customer platforms: 7 of 48 accounts (a04, a11, a12, a17, a20, a24, a35); other maps call this crm.lookup or crm.contact-lookup
The same as data: /api/v1/use-cases/crm.find-record.json, or the MCP tool get_use_case with id crm.find-record. Other jobs in this category: /api/v1/niches?category=crm.
For companies
Do agents find, use and choose your tools?
An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.