HunterAlphaHub
OpenRouter model reference Facts from the public catalogue, dated and labelled
Jev Laya Fable 5.2 Verified
2026-11-13

GitHub · Research and data · Coding and context

kitfunso/hippo-memory

The most carefully reported Jev result in this column, negative included.

The repository's GitHub card, read on 2026-09-23

What it does

A memory layer for AI agents modelled on the hippocampus: decay by default, strength through use, provenance on every memory, SQLite underneath and no runtime dependencies. It works across CLI agents rather than inside one.

The repository's own description: “Biologically-inspired memory for AI agents. Decay, retrieval strengthening, consolidation. Zero runtime deps, SQLite, MCP. Benchmarked retrieval with an opt-in TypeSafe Jev reranker.”

How it works

Jev is an opt-in reranker, and the repository reports it the way a result should be reported: `hippo recall --reranker jev` lifts R@1 from 0.41 to 0.62 on a private 300-query developer store at about $0.0004 a recall, the margin held in 20 of 20 seeds — and three graded tests still did not beat the free local cross-encoder, which is printed in the same document.

The repository, by the numbers

Stars
761
Forks
44
Language
TypeScript
Last push
2026-09-27

Read from the GitHub API on 2026-09-27. Stars and the last push move daily — quote them with the date, the way we do.

What we checked

  • the tree: bin, benchmarks, evals, extensions, integrations, python, docs
  • MIT LICENSE (Keith So) and an npm package name in package.json
  • README reports the Jev reranker eval with its paired-bootstrap margin and its negative result
  • README states the reranker is off by default

What we did not check. We did not install it, and the eval is on a private query store, so it cannot be reproduced from here.

Open the repository ↗All builds

Send us a link

A project built with Jev, a post, a video, a correction, a tip. We open the link, check it says what you said it says, and write the entry ourselves.

Required a link, and an email to reply to. Optional everything else.

Add context — all optional

One or two sentences about what it does, in your words. We write the entry ourselves.

Cost, latency, a benchmark — anything you measured. We attribute these to you.

We store what you type and email it to ourselves. No IP address, no user agent, no referrer — the same rule as the mailing list.

What happens to what you send →