Thread

Jevmem – automatic project memory for Claude Code, built on Jev

62 points · 49 comments · avinashjetwani

rhgraysonii · 1790353996 · focus · HN ↗
Curious how this compares to my own tool, <a href="https://deciduous.dev" rel="nofollow">https://deciduous.dev
I will have to give it a run-through today.
I haven't used Jev yet so this should be interesting. I'd be interested to see if any Deciduous users have opinions, too.
1. avinashjetwani · 1790356866 · focus · HN ↗
  Different shapes, complementary more than competing. jevmem is a thin automatic layer: watch the coding session, keep a flat scored JEVMEM.md (decisions/constraints/bugs/todos) in the repo, inject the relevant live lines next turn. Built around Claude Code hooks; Cursor/Codex share the same file.
  From what I can see on deciduous.dev, you’re building a shared decision graph (goals → options → decisions → observations, multi-agent credit, archaeology from commits, Postgres + MCP). That’s richer reasoning-trail / multi-agent collaboration than jevmem aims for.
  I’d love a side-by-side if you try it — npm i -g jevmem and <a href="https://github.com/Avinash-jetwani/jevmem" rel="nofollow">https://github.com/Avinash-jetwani/jevmem. Curious where deciduous feels clearly better.
2. 0x457 · 1790360370 · focus · HN ↗
  deciduous seems like a great way to ensure there is no space left in context for useful work.
  Problem with all these automatic memory managment tools is that they can't reliably classify if something is useful or not. Once it start using some tool it will heavily become biased towards using that tool during that session. First you will have decision to use postgres as a database, by the end of session it will start saving "decided to use SELECT query for this method because that is how you pull data out of database".
  1. avinashjetwani · 1790362427 · focus · HN ↗
    That’s the hard problem, and I don’t pretend it’s solved. jevmem’s bet is a save/skip gate (Jev probabilities + fixed thresholds in config), not "log every turn." On a held-out set of 66 messages I wrote, save/skip was 98.5%; that is a single author-built run, and I have not measured long-run drift or whether injected lines actually improve answers. Trivial "used SELECT" lines should fail the content/importance gates; if they don’t in real use, that’s a bug I want labeled with `jevmem wrong` / `missed` so `fit` can retune. Happy to hear where deciduous draws that line too.
  2. aifrontierpost · 1790362914 · focus · HN ↗
    [dead]
  3. rhgraysonii · 1790363790 · focus · HN ↗
    Not how it goes in practice at all, would encourage you to actually try it out.
3. avinashjetwani · 1790362307 · focus · HN ↗
  Nice. I have not done a fair side-by-side with Deciduous yet, so I will not invent one. jevmem’s bet is automatic capture into a repo file via Jev’s typed questions (save/skip, kind, contradictions), with Claude Code hooks doing it every turn. Would love Deciduous users’ notes after you try it. Install if useful: npm i -g jevmem · <a href="https://github.com/Avinash-jetwani/jevmem" rel="nofollow">https://github.com/Avinash-jetwani/jevmem

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.