frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

I built a persistent memory layer for AI agents in Rust

1•architsingh15•2h ago
Every Claude Code session starts from zero. It doesn't remember the bug you debugged yesterday, the architecture decision you made last week, or that you prefer Tailwind over Bootstrap. I built Memori to fix this.

It's a Rust core with a Python CLI. One SQLite file stores everything -- text, 384-dim vector embeddings, JSON metadata, access tracking. No API keys, no cloud, no external vector DB.

What makes it different from Mem0/Engram/agent-recall:

- Hybrid search: FTS5 full-text + cosine vector search, fused with Reciprocal Rank Fusion. Text queries auto-vectorize -- no manual --vector flag needed.

- Auto-dedup: cosine similarity > 0.92 between same-type memories triggers an update instead of a new insert. Your agent can store aggressively without worrying about duplicates.

- Decay scoring: logarithmic access boost + exponential time decay (~69 day half-life). Frequently-used memories surface first; stale ones fade.

- Built-in embeddings: fastembed AllMiniLM-L6-V2 ships with the binary. No OpenAI calls.

- One-step setup: `memori setup` injects a behavioral snippet into ~/.claude/CLAUDE.md that teaches the agent when to store, search, and self-maintain its own memory.

Performance (Apple M4 Pro): - UUID get: 43µs - FTS5 text search: 65µs (1K memories) to 7.5ms (500K) - Hybrid search: 1.1ms (1K) to 913ms (500K) - Storage: 4.3 KB/memory, 8,100 writes/sec - Insert + auto-embed: 18ms end-to-end

The vector search is brute-force (adequate to ~100K), deliberately isolated in one function for drop-in HNSW replacement when someone needs it.

After setup, Claude Code autonomously:

- Recalls relevant debugging lessons before investigating bugs

- Stores architecture insights that save the next session 10+ minutes of reading

- Remembers your tool preferences and workflow choices

- Cleans up stale memories and backfills embeddings ~195 tests (Rust integration + Python API + CLI subprocess), all real SQLite, no mocking.

MIT licensed.

GitHub: https://github.com/archit15singh/memori

Blog post on the design principles: https://archit15singh.github.io/posts/2026-02-28-designing-cli-tools-for-ai-agents/

Associated Press Announces It's Teaming Up with Kalshi Ahead of the Midterms

https://www.mediaite.com/media/news/ap-announces-its-teaming-up-with-prediction-market-site-kalsh...
1•terminalbraid•1m ago•0 comments

Why (and how) we built 3 AI agents into our product

https://getlago.com/blog/building-ai-is-hard
1•FinnLobsien•2m ago•0 comments

February 2026: Bitcoin fell 24%. Nothing in crypto infrastructure broke

https://thefutureofmoney.substack.com/p/the-monetary-blueprint-6-cryptos
1•futureofmoney•3m ago•0 comments

YCombodogpatchrental

1•gekahng•5m ago•0 comments

Only 526 AI tools are in the topM most-visited websites

https://airankings.co
1•chadlad101•5m ago•1 comments

Ask HN: How do you get better at coding agents?

1•twitchard•6m ago•0 comments

This Month in Ladybird – February 2026: Adopting Rust for LibJS

https://ladybird.org/newsletter/2026-02-28/
1•exploraz•6m ago•0 comments

Show HN: Vanilla JavaScript refinery simulator built to explain job to my kids

1•fuelingcurious•7m ago•0 comments

What keeps IoT devices running for a decade

https://mooracle.io/blog/building-iiot-that-lasts/
1•mooracle•7m ago•1 comments

Show HN: Synapse – P2P AI agent collaboration with async human supervision

https://github.com/francisdu53/synapse-protocol
1•Fran6codu53•8m ago•0 comments

AI-generated code, AI-generated findings, and the verification bottleneck

https://srlabs.de/blog/ai-verification-bottleneck
3•hacka22•9m ago•0 comments

Show HN: I used an IoT sensor and Claude to diagnose a hairdryer

https://lgdimaggio.github.io/claude-stwinbox-diagnostics/
1•lgdimaggio•11m ago•1 comments

Show HN: Agd – a content-addressed DAG for tracking what AI agents do

https://github.com/frontman-ai/agd
1•BlueHotDog2•12m ago•0 comments

Anthropic to Department of Defense: Drop Dead

https://www.computerworld.com/article/4138860/anthropic-to-department-of-defense-drop-dead.html
1•CrankyBear•13m ago•0 comments

Abstraction Beats Realism in VR Concert Recreation

https://kaikunze.de/post/2026-02-26-abstraction-beats-realism/
1•kgarten•13m ago•0 comments

Kanban Code - Native MacOS UI for Managing Multiple Claude Codes

https://github.com/langwatch/kanban-code
1•jangletown•14m ago•0 comments

Show HN: PRD Agent – Turn your app idea into a shareable RPD in minutes

https://www.prdagent.ai
1•kmatthews812•14m ago•0 comments

We Claudified our iOS app without wrecking our codebase

https://www.tolans.com/relay/how-we-claudified-our-ios-app-without-wrecking-our-codebase
4•quintendf•15m ago•0 comments

Home Assistant can run DOOM

https://frenck.dev/home-assistant-can-run-doom/
1•skortekaas•18m ago•0 comments

California introduces age verification law for all operating systems

https://www.tomshardware.com/software/operating-systems/california-introduces-age-verification-law
3•Alupis•18m ago•1 comments

Israel spent years hacking Tehran’s traffic cameras and monitoring bodyguards

https://www.ft.com/content/bf998c69-ab46-4fa3-aae4-8f18f7387836
1•ent101•19m ago•0 comments

End of hallucinations? How Vancouver AI firms achieve accuracy

https://www.biv.com/news/technology/end-of-hallucinations-how-vancouver-ai-firms-achieve-accuracy...
1•ClearwayLaw•20m ago•1 comments

How to Recover Your Stolen Crypto After a Scam–Guidance from Intelligence Cyber

1•Petersrobert•22m ago•0 comments

CSOA Forte Prenestino (2023)

https://thecommoner.org/csoa-forte-prenestino/
1•greyb•22m ago•0 comments

Prohibited Countries – Mercury Bank

https://support.mercury.com/hc/en-us/articles/28771710754580-Prohibited-countries
1•absqueued•23m ago•1 comments

API to Clean Markdown Docs for AI Agents (No More Stale Endpoints)

https://www.apiflora.dev/
1•LuigiR1•25m ago•1 comments

Dr Seuss Day: 'Without Oxford University, We Don't Get Dr. Seuss'

https://www.bbc.com/news/articles/clywx08zqevo
1•1659447091•25m ago•0 comments

Connected Claude to a 1983 oscilloscope [video]

https://www.youtube.com/shorts/MJHImAx0dAc
1•kmikeym•27m ago•1 comments

FFmpeg at Meta: Media Processing at Scale

https://engineering.fb.com/2026/03/02/video-engineering/ffmpeg-at-meta-media-processing-at-scale/
2•root670•27m ago•0 comments

Managed OpenClaw hosting your own AI assistant in 60 seconds, no server needed

https://www.myopenclaw.cloud
1•danielthego•28m ago•1 comments