Skip to main content
POST
Not everything worth knowing lives in your source systems. A memory lets you tell Seyn something directly, in plain language, and have it processed into structured knowledge. Send a sentence or short paragraph (“deals over $50M always need board sign-off, not just the CEO”) and Seyn extracts the claims, reconciles them against what it already believes, and writes them as assertions attributed to a human edit. A memory can introduce new knowledge or supersede an existing rule that’s now wrong.
  • Asynchronous. The call returns 202 with a memoryId. Processing happens in the background; poll Get memory status to see what Seyn understood and which assertions it created or superseded.
  • Targeted. Memories write into the active library unless you set libraryId.
  • Scoped. Requires an API key with the ingest scope.
This is the same propose-then-apply path the dashboard uses when a human corrects a rule, so every memory leaves an audited trail.

Authorizations

​
Authorization
string
header
required

Bearer token from app.seynlabs.com → Settings → API Keys. Send as Authorization: Bearer sk_live_....

Headers

​
Idempotency-Key
string<uuid>

Optional. Reuse the same key when retrying so a memory is processed only once.

Body

application/json
​
text
string
required

Plain-language description of what you want Seyn to know. A sentence or short paragraph works best.

Required string length: 1 - 4000
Example:

"Deals over $50M always need board sign-off, not just the CEO."

​
libraryId
string<uuid>

Library to write into. Defaults to the active library.

Response

Memory accepted for processing.

​
success
enum<boolean>
required
Available options:
true
​
data
object
required
​
meta
object
required