开发路线图
最早是一个执行很快的规则引擎，现在的方向是业务规则层：规则可以由人或 AI 来写，Ordo 负责测试和检查影响。
Phase 1: Foundation
v0.1 → v0.2 已完成- Core Rule Engine (Rust)
- HTTP & gRPC API Server
- Visual Rule Editor (Vue 3 + React)
- WASM Browser Execution
- Execution Tracing & Debugging
- JIT Compilation (Cranelift)
- Audit Logging & Version History
- Distributed Deployment (NATS JetStream)
Phase 2: Intelligence
v0.3 → v0.4 已完成- Data Filter API (SQL / JSON / MongoDB push-down)
- Partial Evaluation Engine
- Compiled Ruleset (.ordo binary format)
- Multi-tenant QPS Isolation
- Enterprise Plugin System
Phase 3: Rules people and AI can write
Shipped 已完成- Ordo Guard: deterministic guardrails for coding agents
- Exact decimal type for money
- Declared input schema (invalid input is rejected)
- Expression strings in assignments and outputs
- Engine-native decision tables (first / collect hit policies)
- ordo impact: see which decisions an edit changes
- MCP server so agents can test and check their own edits
Phase 4: Ecosystem
Next 进行中- Reusable rule packs (credit approval, promo stacking, fraud scoring)
- Decision table steps in the Studio visual editor
- Sharing rule packs between teams
- Published Go / Python / Java SDKs
- Collaborative editing
- Kubernetes native operator