freshcrate
Skin:/
Home > RAG & Memory > latticesql

latticesql

Persistent structured memory for AI agent systems — pluggable SQLite or Postgres backend, LLM context bridge

Why this rank:Recent releaseHealthy release cadence

Description

Persistent structured memory for AI agent systems — pluggable SQLite or Postgres backend, LLM context bridge

Release History

VersionChangesUrgencyDate
v5.0.0Latest release: v5.0.0High7/20/2026
v4.3.8Latest release: v4.3.8Medium6/26/2026
v4.2.4Latest release: v4.2.4High6/24/2026
v3.0.0Clouds are now a shared Postgres database secured by real Postgres Row-Level Security — each member connects as their own scoped, non-superuser role; the server/replica/sync model is removed (BREAKING). Highlights: - RLS cloud model: private-by-default rows, owner-controlled sharing (private/everyone/custom), scoped member roles. - Per-column audiences (cell masking) + per-viewer enrichment (observe/foldForViewer) with crypto-shred. - S3-backed file bytes for cloud workspaces (opt-in, content-aHigh6/14/2026
v1.13.5## Fixed - **`redeemInvite` HTTP 404 against Postgres cloud URLs.** Same architectural pattern as the v1.13.3 register fix — `fetch()` against a `postgres://...` URL can't reach `/api/auth/redeem-invite`. New `redeemInviteDirect()` in `src/teams/direct-ops.ts` runs the same INSERT sequence directly against the cloud Postgres (token hash compare, email binding, expiry, un-redeemed-yet check). `TeamsClient.redeemInvite` dispatches on URL scheme. - **Create-team / Join-team modal copy.** Cloud URLHigh5/26/2026
main@2026-05-25Latest activity on main branchHigh5/25/2026
1.7.0Imported from npm (1.7.0)Low4/21/2026
main@2026-04-21Latest activity on main branchHigh4/21/2026
v1.0.0Tag v1.0.0Medium4/4/2026
v0.18.4Tag v0.18.4Medium4/4/2026
v0.18.3Tag v0.18.3Medium4/3/2026
v0.18.2Tag v0.18.2Medium4/3/2026
v0.18.1Tag v0.18.1Medium4/3/2026
v0.18.0Tag v0.18.0Medium4/3/2026
v0.16.1Tag v0.16.1Medium4/1/2026
v0.16.0Tag v0.16.0Medium4/1/2026

Dependencies & License Audit

Loading dependencies...

Similar Packages

contextgitPersistent memory layer for AI agent workflowsmain@2026-06-03
@runcore-sh/runcoreLocal-first AI agent runtime. One command, your machine, your data.0.6.0
@matanelcohen/openspace-aiVisual workspace manager for Squad SDK — create, manage, and collaborate with AI agent teams0.5.0
contextmateZero-knowledge encrypted sync for AI agent context0.4.29
universal-agent-memoryAutonomous AI agent memory system with CLAUDE.md protocol enforcementv1.172.1

More in RAG & Memory

edgequakeEdegQuake 🌋 High-performance GraphRAG inspired from LightRag written in Rust; Transform documents into intelligent knowledge graphs for superior retrieval and generation
vllmA high-throughput and memory-efficient inference and serving engine for LLMs
awesome-opensource-aiCurated list of the best truly open-source AI projects, models, tools, and infrastructure.
rust-sdkThe official Rust SDK for the Model Context Protocol