Skip to main content
The retrieval counterpart to save_note. Semantically searches the meeting briefs, transcripts, meeting notes, and notes kept on contacts, and returns the matching documents with a snippet — so the agent pulls the relevant passage rather than loading whole documents into context. Pass focus to restrict to one person or company, or omit it to search across everyone in the workspace.

​
Parameters

NameTypeRequiredDescription
questionstring✓Natural-language query matched against document content
focusstringRestrict to one person/company — email, LinkedIn URL, domain, or entity UUID. Omit to search everyone.
limitnumberMax documents to return. Default 8, cap 20.

​
Returns

Ranked documents, each with type, title, date, similarity, and a snippet. Get the full body with get_account. 
Documents matching "what did we discuss about pricing":

  meeting notes · Q3 renewal call  (88%)  [2d ago]
    Sarah pushed back on per-seat — wants a flat platform fee. Maria (CFO) to…
    (entity_id: a1b2c3d4-...)
  meeting brief · Pre-meeting brief — renewal  (74%)  [1w ago]
    Budget waiting on Q3 close; proposal at $12k/yr, expect a discount ask…
    (entity_id: a1b2c3d4-...)
If nothing matches: 
No saved documents matched "what did we discuss about pricing".

​
When to call it

  • “What did we discuss about pricing?” — pull the relevant past notes
  • “What objections came up in past meetings?” — across a contact’s record
  • Comparing across meetings — restrict with focus and read the snippets

​
Backed by

POST /v2/notes/search — see the HTTP reference for the raw JSON shape.