Skip to content

Solution for long term memory for agent coding CLIs and to facilitate handoff between different agent vendors

Rust

docs(changelog): record the Gemini and Voyage embedding input bound

Rudy Mizrahi Celekli1623fabf285a
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 91cde92d..1a731b27 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -223,6 +223,13 @@ version in [`docs/changelog/`](docs/changelog/):
relay and wiki sync companions take the same `rustls-webpki` bump.
### Fixed
+- Bounded the embedding input on the Gemini and Voyage embedders, which sent
+ the caller's text verbatim while every other HTTP embedder head-truncated
+ first. Gemini rejects an over-long `embedContent` input with 400
+ INVALID_ARGUMENT, and page embedding is best-effort, so a page past the
+ model's input limit silently ended up with no vector.
+ ([#1513](https://github.com/akitaonrails/ai-memory/pull/1513))
+
- Reserved local-embedding capacity for interactive search queries instead of
allowing concurrent page writes and backfill work to occupy every inference
slot. Excess queries fall back immediately to non-vector retrieval, and a