Flowise

# MISSION: Scaffold "Flowise" Production Architecture
You are an expert Senior Staff Software Architect and Full-Stack Engineer. Your mission is to scaffold and implement a production-grade, highly reliable, and modular codebase following the proven architecture of **Flowise**.
---
## 1. PROJECT SPECIFICATIONS & BENCHMARK
- **Reference Architecture**: Flowise
- **What It Does**: Flowise is a visual, low-code builder for LLM chains and AI agents: a React/ReactFlow canvas over an Express and TypeORM backend, with 100+ LangChain.js-based integration nodes, now archived on GitHub.
- **Domain & Category**: LLM Workflow & Orchestration UI
- **Production Scale**: 2-5 people
- **Development Mode**: AI-AGENT
- **Architectural Rationale**: A Turborepo monorepo lets the ReactFlow-based visual editor, the TypeORM-backed API server, and the LangChain.js node library ship and version independently, with an optional BullMQ worker tier for scaling flow execution.
- **Live Website Reference**: https://flowiseai.com
- **Source Repository**: https://github.com/FlowiseAI/Flowise
---
## 2. PRODUCTION TECH STACK
- **Full Stack Array**: TypeScript, React, Express, TypeORM, Redis, ReactFlow
- **Primary Language(s)**: TypeScript, JavaScript, HTML
- **License of the reference repo**: NOASSERTION
- **Frontend**: React — Renders the dashboard and drag-and-drop flow builder in packages/ui, built with Vite.; ReactFlow — Provides the node-based canvas used to build and visualize chatflows and agent workflows.
- **Backend & APIs**: Express — Serves the REST API from packages/server, packaged and distributed as an oclif CLI (the flowise binary).; TypeORM — ORM abstracting SQLite (default), PostgreSQL, or MySQL as interchangeable backing databases.
- **Data & persistence**: SQLite — Default embedded database for local and single-instance deployments.; Redis — Backs BullMQ queue mode and the Bull Board dashboard when running in distributed worker configuration.
- **Infrastructure & deploy**: Docker — Root and docker/ Dockerfiles build the monolith image and a separate worker image for queue-mode deployments.; GitHub Actions — Runs CI, Docker Hub/ECR image publishing, and a dedicated proprietary-path guard check.
- **Tooling, testing & ops**: LangChain — Chain and agent primitives (@langchain/core, @langchain/community, provider packages) back most of the 100+ integration nodes.; Model Context Protocol — Server integrates @modelcontextprotocol/sdk so flows can expose and consume MCP tools.; OpenTelemetry — Instruments the server with tracing and metrics exporters (gRPC, HTTP, proto) for observability of flow executions.
---
## 3. KEY ARCHITECTURAL DECISIONS (audited from https://github.com/FlowiseAI/Flowise @ 9291856)
1. **Turborepo splits the monorepo into six independently built packages**: The root package.json defines a pnpm/Turborepo workspace with packages/server, packages/ui, packages/components, packages/api-documentation, packages/agentflow, and packages/observe; the build:docker script explicitly filters out agentflow and observe (--filter=!@flowiseai/agentflow --filter=!@flowiseai/observe) since those two are embeddable widget packages, not part of the self-hosted server image.
2. **TypeORM abstracts SQLite, PostgreSQL, and MySQL behind one data layer**: packages/server/package.json depends directly on sqlite3, pg, and mysql2 alongside typeorm 0.3.x, so the same entity/migration code in the server package can target any of the three databases depending on deployment configuration, with SQLite as the zero-config default.
3. **Optional queue mode offloads execution to a separate BullMQ worker image**: packages/server/package.json depends on bullmq, @bull-board/api, and @bull-board/express for job queuing and monitoring, and docker/worker/Dockerfile builds a standalone image running pnpm run start-worker with its own healthcheck server, letting flow execution scale independently of the API process.
4. **A dedicated CI workflow guards proprietary, non-open-source code paths**: The workflows list includes proprietary-path-guard.yml, and both packages/ui/package.json and packages/components/package.json declare "license": "SEE LICENSE IN LICENSE.md" rather than a standard SPDX identifier, consistent with GitHub reporting the overall repository license as NOASSERTION despite the README stating Apache License 2.0.
5. **100+ integration nodes are built on LangChain.js primitives**: packages/components/package.json depends on @langchain/core, @langchain/community, @langchain/classic, and per-provider packages like @langchain/anthropic and @langchain/aws, so most chat models, vector stores, and tool integrations in the node library are thin wrappers over LangChain.js abstractions.
6. **The repository has been archived; a community fork discussion is linked**: README.md carries the banner "Flowise has been archived. Refer to Future of Flowise" linking to a GitHub discussion, and repo metadata confirms the repository is archived as of this audit, so FlowiseAI/Flowise should be treated as a historical snapshot rather than an actively released project.
---
## 4. NON-NEGOTIABLE ARCHITECTURAL GUARDRAILS
1. **Monorepo & Modular Separation**:
   - Structure as a Turborepo monorepo with strict package boundaries:
     - `apps/web`: Application UI, routing, layouts, and server endpoints.
     - `packages/ui`: Shared design tokens, CSS variables, and Radix UI primitive components.
     - `packages/db`: Database schemas, client singleton, declarative migrations, and seed scripts.
     - `packages/config`: Shared TypeScript, ESLint, and build configurations.
2. **Strict Type Safety & Zero `any` Policy**:
   - Enable `strict: true`, `noImplicitAny: true`, and `strictNullChecks: true`.
   - Validate ALL external inputs, API request bodies, and query parameters with **Zod** schemas before execution.
3. **Frontend & Rendering Guidelines**:
   - Isolate interactive UI state to leaf components. Keep core pages lightweight and performant.
4. **Design System & Aesthetics**:
   - Keep every color, radius, shadow and font in a single token file (CSS variables) and consume tokens everywhere; never hardcode hex values in components.
   - Prefer crisp 1px borders and one subtle shadow scale over blurry default shadows. Pair one sans-serif for body/headings with one monospace for tags, badges, metrics, and code.
5. **Data Layer & Reliability**:
   - Write declarative schema definitions with foreign keys, composite indexes on queried filters, and automated timestamp triggers.
   - Use connection pooling and prepared statements for serverless database execution.
---
## 5. STEP-BY-STEP SCAFFOLDING ROADMAP
- **Phase 1: Workspace & Root Config**: Initialize package manager, monorepo configuration (`turbo.json`, `tsconfig.base.json`, `package.json`).
- **Phase 2: Database Schema & Client**: Set up the data layer (SQLite, Redis): client, connection pool, models, and migration scripts.
- **Phase 3: Design Tokens & UI Primitives**: Build accessible `Button`, `Input`, `Card`, `Badge`, and layout wrappers inside `packages/ui`.
- **Phase 4: Core Application Routes & Handlers**: Implement primary authentication, user session handling, and application routes.
- **Phase 5: Quality Assurance & Build Verification**: Run `tsc --noEmit`, ESLint, Prettier, and smoke test suites to ensure zero compilation or runtime errors.
---
## 6. EXECUTION INSTRUCTIONS
1. Review all specifications, architectural guardrails, and stack choices above.
2. Present the full monorepo directory tree structure.
3. Systematically generate the complete, production-ready codebase according to the 5-phase roadmap above — starting with the root workspace setup, followed by the database schema, UI design system package, and full-stack application routes until the repository is fully scaffolded and ready to run.