Local-First | Lab | E2EE | Infra | Benchmarks | Tools | CLI | Adapters | CRDT Sync | Relays
Public rollout is in progress (updated March 1, 2026).
- 🌍 Grow the local-first community and make the ecosystem easier to discover and adopt.
- 🛠️ Build practical tools and a robust CLI for developers and operators.
- ⚡ Push performance in critical paths with Zig/Rust and systems-level optimizations.
- 🧩 Keep Evolu compatibility while delivering upgrades and upstream-ready improvements.
- 🧪 Run experiments, publish benchmarks, and propose new local-first specs where needed.
- 🤖 Advance local-first collaboration patterns for AI agents.
SQLoot currently runs on top of an Evolu-based baseline. Our immediate strategy is to preserve upstream compatibility, ship measurable improvements in our fork, and upstream changes where applicable.
- Upstream project: evoluhq/evolu
- Upstream docs: evolu.dev
- SQLoot baseline fork: SQLoot/evolu-plan-b
Credit where it belongs: huge respect to evoluhq and @steida for creating Evolu and moving the local-first ecosystem forward.
| Track | Role | Current State |
|---|---|---|
evolu-plan-b |
Evolu-compatible Bun + Biome baseline | Public rollout |
core |
SQLoot-native runtime foundations (Rust/Zig) | Planning |
bench-suite |
Reproducible benchmark harness | WIP |
relays |
Evolu-compatible relay implementations (Bun+Elysia/Rust/BEAM) | WIP |
lootbox |
CLI umbrella, TUI, tools (Rust) | WIP |
blind-cloud |
Encrypted cloud transport experiments (vault/relay/repacker) | Research |
local-first |
Community forum, blog, wiki, guides | TBD |
loot-cms |
File-based CMS, editor, sync storage | Research |
loot-ops |
Org-wide platform glue for repos (Nix/IaC/CI) | Private |
website |
Landing page, docs, bench measurements, stats | WIP |
knowledge |
Obsidian KB, planning, issues, research | Internal |
Public docs and public APIs are opening in phases as each track stabilizes.
- 📦 Public track for
evolu-plan-bprepared (100%) - 🧩 Evolu-based Bun-first baseline established (95%)
- 🔌 Priority integrations (TanStack Start, Tauri, Astro) (75%)
- 🧪 Benchmark harness foundation in
bench-suite(45%) - 📊 Regression evidence gates (p50/p95/throughput/memory) (40%)
- 🌐 Relay and transport hardening (
relays,blind-cloud) (30%) - 🧱 SQLoot core bootstrap (
core) (20%) - 🛠️
lootboxCLI baseline (scope, commands, packaging) (25%) - 📚 Public docs and evidence surfaces (
website,local-first) (35%)
- Lock
evolu-plan-bpublic baseline: keep CI green, tighten docs, and close rollout gaps. - Ship
bench-suitebaseline lanes with reproducible report schema and first regression thresholds. - Stabilize integration lanes for TanStack Start, Tauri, and Astro with actionable smoke outputs.
- Define and run relay hardening matrix for reconnect/load/failure handling in
relays. - Ship
lootboxCLI v0 scope and first operator commands for workspace/bootstrap + benchmark flows.
Contributions are welcome. See CONTRIBUTING for workflow and quality standards.
For responsible disclosure, see SECURITY.
