Skip to content
/ workflow-cccc Public

基于cccc-sdk 理论上创建一切可执行的工作流

License

MIT license
2 stars 0 forks Branches Tags Activity
Star
Notifications You must be signed in to change notification settings

dweb-channel/workflow-cccc

BranchesTags

Repository files navigation

Work-Flow

English | 中文

English

An open-source developer workflow platform that automates repetitive engineering tasks using Claude AI + Temporal + LangGraph.

What It Does

Two production pipelines:

Pipeline Input Output How It Works
Batch Bug Fix Jira URLs Git PRs with fixes Jira fetch → Claude CLI analysis → code fix → verification → Git branch + PR
Design-to-Spec Figma URL design_spec.json Figma API scan → frame classification → two-pass SpecAnalyzer (LLM vision) → structured spec

Both pipelines run as durable Temporal workflows with real-time SSE progress streaming.

Tech Stack

Layer Technology
Frontend Next.js 14, React Flow, shadcn/ui, Tailwind CSS
Backend FastAPI, SQLAlchemy (async), Pydantic v2
Orchestration Temporal (durable workflows), LangGraph (DAG execution)
AI Claude CLI (code generation), Claude API (vision analysis)
Database SQLite (aiosqlite) — PostgreSQL ready
Testing pytest (624 tests), Vitest, Playwright

Quick Start

Option A: Local Development

Prerequisites: Python 3.11+, Node.js 18+, Temporal CLI

# Clone
git clone https://github.com/your-org/work-flow.git
cd work-flow

# Backend
cd backend
python -m venv .venv
source .venv/bin/activate
pip install -e ".[test]"
make dev    # Starts Temporal + Worker + FastAPI on :8000

# Frontend (new terminal)
cd frontend
npm install
npm run dev  # Starts Next.js on :3000

Option B: Docker Compose

cp .env.example .env
# Edit .env — set FIGMA_TOKEN, JIRA_* if needed
docker compose up --build

Services: backend :8000, frontend :3000, Temporal :7233 (Web UI :8080)

Open http://localhost:3000

Configuration

Copy .env.example to .env and configure:

Variable Required Description
ANTHROPIC_API_KEY For AI features Claude API key
FIGMA_TOKEN For Design-to-Spec Figma Personal Access Token
JIRA_URL For Batch Bug Fix Jira instance URL
JIRA_EMAIL For Batch Bug Fix Jira account email
JIRA_API_TOKEN For Batch Bug Fix Jira API token
CLAUDE_CLI_PATH Auto-detected Path to claude CLI binary

Project Structure

work-flow/
├── backend/
│   ├── app/                    # FastAPI application
│   │   ├── routes/             # API endpoints (batch, design, workflows)
│   │   ├── repositories/       # Data access layer
│   │   ├── models/             # SQLAlchemy ORM models
│   │   ├── database.py         # DB engine + session management
│   │   ├── event_bus.py        # Unified SSE event infrastructure
│   │   └── temporal_adapter.py # Temporal client singleton
│   ├── workflow/
│   │   ├── engine/             # LangGraph graph builder + executor
│   │   ├── nodes/              # Node type registry (agents, spec_analyzer, etc.)
│   │   ├── temporal/           # Temporal workflows + activities
│   │   ├── integrations/       # Figma client, classifiers
│   │   └── templates/          # Workflow JSON templates
│   ├── tests/                  # 624 tests (pytest)
│   ├── Makefile                # dev/stop/worker/api/temporal
│   └── pyproject.toml
├── frontend/
│   ├── app/
│   │   ├── batch-bugs/         # Batch Bug Fix UI
│   │   ├── design-to-code/     # Design-to-Spec UI
│   │   └── workflows/          # Visual workflow editor
│   ├── components/             # Shared UI components (shadcn/ui)
│   └── lib/                    # Hooks, utilities
├── docker-compose.yml          # 3-service stack
├── Dockerfile.backend          # Multi-stage Python build
├── Dockerfile.frontend         # Next.js standalone build
└── .env.example                # Environment template

API Endpoints

Batch Bug Fix

Method Path Description
POST /api/v2/batch/bug-fix Create batch job (or dry-run with dry_run: true)
GET /api/v2/batch/bug-fix List jobs (pagination, status filter)
GET /api/v2/batch/bug-fix/{job_id} Job status + bug details
GET /api/v2/batch/bug-fix/{job_id}/stream SSE event stream
POST /api/v2/batch/bug-fix/{job_id}/cancel Cancel running job
POST /api/v2/batch/bug-fix/{job_id}/retry/{bug_index} Retry failed bug
DELETE /api/v2/batch/bug-fix/{job_id} Delete job
GET /api/v2/batch/metrics/job/{job_id} Job metrics
GET /api/v2/batch/metrics/global Global metrics

Design-to-Spec

Method Path Description
POST /api/v2/design/scan-figma Scan Figma page, classify frames
POST /api/v2/design/run-spec Start spec generation pipeline
GET /api/v2/design/{job_id} Job status
GET /api/v2/design/{job_id}/stream SSE event stream
GET /api/v2/design/{job_id}/files Generated spec files
GET /api/v2/design List jobs
POST /api/v2/design/{job_id}/cancel Cancel job

Visual Workflows

Method Path Description
GET /api/v2/workflows List workflows
POST /api/v2/workflows Create workflow
POST /api/v2/workflows/{id}/run Execute workflow
GET /api/v2/sse/stream/{run_id} SSE event stream

Testing

cd backend

# All tests
python -m pytest tests/ -q

# By module
python -m pytest tests/test_routes_batch.py -q       # 34 route tests
python -m pytest tests/test_routes_design.py -q       # 22 route tests
python -m pytest tests/test_repository_batch_job.py -q # 28 repo tests
python -m pytest tests/workflow/ -q                    # Workflow engine tests

Architecture

[Browser] ──→ [Next.js :3000] ──→ [FastAPI :8000] ──→ [Temporal :7233]
                                        │                     │
                                   [SQLite DB]          [Worker Process]
                                        │                     │
                                   [EventBus] ←── SSE ──── [Activities]
                                        │                     │
                                   [SSE Stream] ←──────── [Claude CLI / Figma API]

中文

开源的开发者工作流平台，使用 Claude AI + Temporal + LangGraph 自动化重复性工程任务。

核心功能

两条生产管线：

管线 输入 输出 工作方式
批量修 Bug Jira URL 列表 Git PR（含修复） Jira 获取 → Claude CLI 分析 → 代码修复 → 验证 → Git 分支 + PR
设计转 Spec Figma URL design_spec.json Figma API 扫描 → 帧分类 → 两轮 SpecAnalyzer（LLM 视觉分析） → 结构化规格

两条管线均以 Temporal 持久化工作流运行，支持实时 SSE 进度推送。

技术栈

层级 技术
前端 Next.js 14, React Flow, shadcn/ui, Tailwind CSS
后端 FastAPI, SQLAlchemy (async), Pydantic v2
编排 Temporal（持久化工作流）, LangGraph（DAG 执行）
AI Claude CLI（代码生成）, Claude API（视觉分析）
数据库 SQLite (aiosqlite) — 可切换 PostgreSQL
测试 pytest（624 个测试）, Vitest, Playwright

快速启动

方式 A：本地开发

前置条件： Python 3.11+、Node.js 18+、Temporal CLI

# 克隆
git clone https://github.com/your-org/work-flow.git
cd work-flow

# 后端
cd backend
python -m venv .venv
source .venv/bin/activate
pip install -e ".[test]"
make dev    # 一键启动 Temporal + Worker + FastAPI（:8000）

# 前端（新开终端）
cd frontend
npm install
npm run dev  # 启动 Next.js（:3000）

方式 B：Docker Compose

cp .env.example .env
# 编辑 .env — 按需设置 FIGMA_TOKEN、JIRA_* 等
docker compose up --build

服务：后端 :8000、前端 :3000、Temporal :7233（Web UI :8080）

打开 http://localhost:3000

配置说明

将 .env.example 复制为 .env 并配置：

变量 是否必需 说明
ANTHROPIC_API_KEY AI 功能需要 Claude API 密钥
FIGMA_TOKEN 设计转 Spec 需要 Figma 个人访问令牌
JIRA_URL 批量修 Bug 需要 Jira 实例地址
JIRA_EMAIL 批量修 Bug 需要 Jira 账号邮箱
JIRA_API_TOKEN 批量修 Bug 需要 Jira API 令牌
CLAUDE_CLI_PATH 自动检测 claude CLI 二进制路径

项目结构

work-flow/
├── backend/
│   ├── app/                    # FastAPI 应用
│   │   ├── routes/             # API 端点（batch, design, workflows）
│   │   ├── repositories/       # 数据访问层
│   │   ├── models/             # SQLAlchemy ORM 模型
│   │   ├── database.py         # 数据库引擎 + 会话管理
│   │   ├── event_bus.py        # 统一 SSE 事件基础设施
│   │   └── temporal_adapter.py # Temporal 客户端单例
│   ├── workflow/
│   │   ├── engine/             # LangGraph 图构建器 + 执行器
│   │   ├── nodes/              # 节点类型注册（agents, spec_analyzer 等）
│   │   ├── temporal/           # Temporal 工作流 + 活动
│   │   ├── integrations/       # Figma 客户端、分类器
│   │   └── templates/          # 工作流 JSON 模板
│   ├── tests/                  # 624 个测试（pytest）
│   ├── Makefile                # dev/stop/worker/api/temporal
│   └── pyproject.toml
├── frontend/
│   ├── app/
│   │   ├── batch-bugs/         # 批量修 Bug 界面
│   │   ├── design-to-code/     # 设计转 Spec 界面
│   │   └── workflows/          # 可视化工作流编辑器
│   ├── components/             # 共享 UI 组件（shadcn/ui）
│   └── lib/                    # Hooks、工具函数
├── docker-compose.yml          # 3 服务编排
├── Dockerfile.backend          # 多阶段 Python 构建
├── Dockerfile.frontend         # Next.js standalone 构建
└── .env.example                # 环境变量模板

测试

cd backend

# 全量测试
python -m pytest tests/ -q

# 按模块
python -m pytest tests/test_routes_batch.py -q       # 34 个路由测试
python -m pytest tests/test_routes_design.py -q       # 22 个路由测试
python -m pytest tests/test_repository_batch_job.py -q # 28 个仓储测试
python -m pytest tests/workflow/ -q                    # 工作流引擎测试

架构

[浏览器] ──→ [Next.js :3000] ──→ [FastAPI :8000] ──→ [Temporal :7233]
                                       │                     │
                                  [SQLite DB]          [Worker 进程]
                                       │                     │
                                  [EventBus] ←── SSE ──── [Activities]
                                       │                     │
                                  [SSE 推送] ←──────── [Claude CLI / Figma API]

License

MIT

About

基于cccc-sdk 理论上创建一切可执行的工作流

Resources

Readme

License

MIT license

Contributing

Contributing

Security policy

Security policy
Activity
Custom properties

Stars

2 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

No packages published

Languages