release: v0.7.0 — event time on ingest, get-by-id, paging - #17
Merged
Conversation
`observe()` accepted an `occurredAt` (documented as "defaults to now server-side")
but dropped it into the opaque `metadata` blob and never set `validFrom` — and
`CreateMemoryRequest` had no `validFrom` field at all, so the SDK was
structurally incapable of back-dating a memory.
`validFrom` is the column behavior mining buckets day-of-week / hour-of-day on,
so a customer backfilling historical orders got patterns describing the moment
they ran the import ("concentrates on Sunday, clusters around 17:00") rather
than when the orders actually happened. Nothing errored — the field validated,
round-tripped on read, and produced plausible-looking output.
Also closes the inspection gaps that same customer hit:
- admin.get(memoryId) — there was no way to resolve a memory id at all.
- admin.listAll() — async-iterator auto-paginator so callers stop hand-rolling
the offset loop.
- search() takes `offset`.
- explain() now point-looks-up its sources instead of scanning
list({ limit: 1000 }), which exceeded the route's max of 500 and 400'd.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Cuts
@thinkfleet/memory-sdk@0.7.0, already published to npm (latest). This lands the release commit onmainso the repo doesn't drift from the registry.The fix
observe()accepted anoccurredAt— documented as "defaults to now server-side" — but dropped it into the opaquemetadatablob and never setvalidFrom.CreateMemoryRequesthad novalidFromfield at all, so the SDK was structurally incapable of back-dating a memory.validFromis the column behavior mining buckets day-of-week / hour-of-day on. A customer backfilling historical orders therefore got every row stamped with the moment of import, and the mined patterns described the import job — "concentrates on Sunday, clusters around 17:00" — rather than their data. Nothing errored: the field validated, round-tripped on read, and produced plausible-looking output.Also
admin.get(memoryId)— there was previously no way to resolve a memory id at all.admin.listAll()— async-iterator auto-paginator.offseton search.explain()now point-looks-up its sources instead of scanninglist({ limit: 1000 }), which exceeded the route's max of 500 and 400'd.npm jumps 0.5.0 → 0.7.0:
mainhad been bumped to 0.6.0 for thebrainsresource, but 0.6.0 was never published. 0.7.0 contains both.🤖 Generated with Claude Code