Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

314 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

yuleOSH

嵌入式软件合规开发自动化平台
Embedded Software Compliance Automation Platform
AI 辅助 · SWE 全流程 · 证据包一键生成

CI Version License Python Tests KG ASPICE R3

pip install yuleosh → running in 2 minutes.
No NDA. No Sales Call. No License Negotiation.

Quick Start · Features · Architecture · Platforms · Pricing · Roadmap


🇬🇧 English · 🇨🇳 中文


📋 Table of Contents


What is yuleOSH?

yuleOSH is an ASPICE assistance tool for embedded development powered by AI. It converts natural language requirements into complete, CI/CD-ready firmware projects with Automotive SPICE traceability assistance.

In one sentence: yuleOSH takes a spec or user story and outputs reviewed, tested, CI-instrumented firmware with ASPICE traceability assistance — all in under 2 minutes.


⚡ Quick Start — 3 Commands

# Step 1: Install (15 seconds)
pip install yuleosh

# Step 2: Initialize a project (15 seconds)
yuleosh init my-project

# Step 3: Run the full pipeline (90 seconds)
cd my-project && yuleosh pipeline run docs/spec.md

3 commands. 2 minutes. From zero to firmware.
No MCU board, no compiler setup, no document reading required.


🎬 Demo — Try the UART Demo

pip install yuleosh
yuleosh demo uart
cd uart-demo && yuleosh pipeline run --mock docs/spec.md

Output you'll see:

Hello from yuleOSH Demo UART
demo UART ready — send characters to echo
[yuleOSH] alive — 5s
[yuleOSH] alive — 10s

Features

🧠 OpenSpec Engine

Structured requirements using RFC 2119 keywords (SHALL/SHOULD/MAY) with GIVEN/WHEN/THEN scenarios. Auto-validates, diffs, and traces every requirement through design → code → test.

🔍 AI Code Review

Parallel 4-agent review matrix covering architecture, domain correctness, coding style, and test coverage. Includes 8 embedded-C static analysis checks plus resource usage prediction (stack, heap, flash, RAM).

🔧 Hardware-in-the-Loop

Built-in adapters for OpenOCD (STM32), JLink (ARM Cortex-M), and esptool (ESP32). Auto-flash, serial monitor, and GDB debugging — one command away.

☁️ SaaS Dashboard

Next.js web dashboard with PostgreSQL multi-tenant storage, JWT authentication, org/project isolation, and real-time pipeline monitoring.

📋 Compliance

One-click generation of traceability matrices, acceptance matrices, and compliance evidence ZIP archives — assists in preparing ASPICE SWE.1~SWE.6 audit evidence.

Full Automation Pipeline

User Story → OpenSpec → SDD → DDD → Code Gen → Internal Review →
Test Planning → Code Review → CI Run → Evidence Pack → Deployment

Architecture

4-Layer Architecture

[User Story / Spec] ──▶ [OpenSpec Engine] ──▶ [Agent Pipeline] ──▶ [Code Gen]
                              │                       │                    │
                              ▼                       ▼                    ▼
                      SHALL/SHOULD/MAY         10-Step Agent        C + Python
                      + GIVEN/WHEN/THEN        Orchestration        Firmware
                                                                         │
                              ┌──────────────────────────────────────────┘
                              ▼
                    ┌─────────────────┐
                    │    Review        │
                    │  (4-Agent Matrix) │
                    └────────┬────────┘
                             ▼
                    ┌─────────────────┐     ┌──────────────────┐     ┌──────────────────┐
                    │   CI Layer 1    │▶──▶ │   CI Layer 2     │▶──▶ │   CI Layer 3     │
                    │  Unit + Coverage │     │  Cross-Compile    │     │  System Verify    │
                    │  + Plan-Lint     │     │  + Static Analysis│     │  + Evidence       │
                    └─────────────────┘     └──────────────────┘     └──────────────────┘
                                                                              │
                                                                              ▼
                                                                     ┌──────────────────┐
                                                                     │  Deploy Hardware │
                                                                     │  OpenOCD / JLink  │
                                                                     │  / esptool        │
                                                                     └──────────────────┘

Layer Details

1. OpenSpec Engine — Spec parsing, validation, version diff, state machine
  • Parser: SHALL/SHOULD/MAY + GIVEN/WHEN/THEN
  • Validator: Hierarchical requirement IDs (SYS/SW/FEATURE)
  • Differ: Version-to-version delta with impact analysis
  • State machine: PROPOSED → APPROVED → IMPLEMENTED → VERIFIED
  • Location: src/yuleosh/spec/
2. Agent Pipeline — 10-step LLM orchestration
  • 10-step orchestration: spec → SDD → DDD → code → test → review
  • LLM-agnostic client (OpenAI-compatible API)
  • Blocking review gates before each stage transition
  • S.U.P.E.R. startup analysis for new requirements
  • Location: src/yuleosh/pipeline/, src/yuleosh/llm/
3. CI/CD Engine — 3-layer automated verification
  • Layer 1 — Dev Verify: Unit tests + coverage gate + plan-lint on every commit
  • Layer 2 — Integration: Cross-compilation + MISRA static analysis on MR
  • Layer 2.5 — AI Review: 4-agent parallel code review
  • Layer 3 — System Verify: System tests + evidence pack on release tag
  • Location: src/yuleosh/ci/
4. Hardware & Cross-Compilation — MCU flashing, monitoring, debugging
  • Target configuration for MCU families
  • Flash, monitor, and debug orchestration
  • SIL (Software-in-the-Loop) runner with assertion checking
  • Extensible adapter architecture
  • Location: src/yuleosh/cross/, src/yuleosh/hardware/

Supporting Modules

Module Path Purpose
Evidence Engine src/yuleosh/evidence/ Traceability matrix + acceptance matrix + compliance ZIP
Review Engine src/yuleosh/review/ 4-agent parallel review + resource predictor
Test Generation src/yuleosh/testgen/ Auto-generate test harness from spec scenarios
Plugins src/yuleosh/plugins/ Plugin registry + sandboxed execution
Usage/Billing src/yuleosh/usage/ Metering + Stripe gateway (for SaaS)
CLI src/yuleosh/cli/ 12+ subcommands
API src/yuleosh/api/ REST API v1 with 14 resource handlers
Dashboard UI frontend/ Next.js web dashboard
Preview src/yuleosh/preview/ Pre-pipeline analysis & scoring

Supported Platforms

Platform Flash Tool Debugger
ESP32 / ESP32-S3 esptool idf-monitor + GDB
STM32 (F4/H7/G0) OpenOCD OpenOCD + GDB
Any ARM Cortex-M JLinkExe JLinkGDBServer
Custom Plugin API Plugin API

Directory Layout

yuleOSH/
├── src/yuleosh/
│   ├── spec/          OpenSpec parser, validator, differ
│   ├── pipeline/      Agent pipeline orchestrator (10 steps)
│   ├── ci/            3-layer CI/CD with dependency chaining
│   ├── review/        4-agent parallel review + resource predictor
│   ├── evidence/      Traceability + acceptance + compliance ZIP
│   ├── hardware/      Flash, monitor, debug orchestration
│   ├── cross/         Cross-compilation + HIL/SIL runners
│   ├── testgen/       Auto test harness generation
│   ├── llm/           LLM-agnostic agent client
│   ├── plugins/       Plugin registry + sandbox
│   ├── api/           REST API v1 (14 handlers)
│   ├── ui/            Dashboard server (auth, routes)
│   ├── cli/           CLI subcommands
│   ├── usage/         Metering + billing integration
│   ├── preview/       Pre-pipeline analysis & scoring
│   └── store.py       Multi-tenant SQLite/PostgreSQL backend
├── frontend/          Next.js SaaS dashboard
├── tests/             250+ tests (all passing)
├── docs/              Specifications, guides, reports
├── deploy/            Production deployment configs
├── Dockerfile         Multi-stage production Dockerfile
├── docker-compose.yml Production Docker Compose
├── install.sh         One-line production install
└── pyproject.toml     Python packaging

Production Deployment

Docker Compose (Recommended)

git clone https://github.com/frisky1985/yuleOSH.git
cd yuleOSH
cp deploy/.env.production.example deploy/.env.production
# Edit deploy/.env.production with your secrets
docker compose -f deploy/docker-compose.yml up -d

pip Install (Standalone CLI)

pip install yuleosh
yuleosh init my-project
yuleosh pipeline run docs/spec.md

One-Line Install (Full Suite)

curl -fsSL https://raw.githubusercontent.com/frisky1985/yuleOSH/main/install.sh | bash

From Source

git clone https://github.com/frisky1985/yuleOSH.git
cd yuleOSH
pip install -e .
yuleosh init .
yuleosh help

Pricing & Editions

yuleOSH offers three editions tailored to different needs. See the full Edition Matrix → for a detailed feature comparison.

Edition Price Best For
Community (MIT) ¥0 Individual developers, open-source projects
SaaS Pro ¥599/mo (¥5,999/yr) Embedded teams needing ASPICE compliance + full pipeline
Enterprise ¥98,000/yr Large organizations needing private deployment + RMB contract support

📖 Full Edition Matrix → — Detailed feature comparison across all editions.


Roadmap

Version Focus Status
v0.1.0 Foundation — OpenSpec, agent pipeline, CI/CD, evidence
v0.2.0 ASPICE compliance — strict mode, bidirectional tracing
v0.3.0 Ground reinforcement — test planning, hierarchy, cross-compile
v1.0.0 Production — HIL adapter, plugin marketplace, scaling 🚧
v1.1.0 Enterprise — RBAC, audit logging, SAML SSO 📋
v1.2.0 Cloud — multi-region, data residency, managed hosting 📋

Contributing

See CONTRIBUTING.md for development setup, code conventions, and PR workflow.

Security

See SECURITY.md for our vulnerability disclosure process.

License

MIT License — see LICENSE for details. Copyright (c) 2025 frisky1985.


Built for embedded teams who ship quality firmware, fast.


yuleOSH — 嵌入式软件合规开发自动化平台

📋 目录


项目简介

yuleOSH 是嵌入式软件合规开发自动化平台(ASPICE SWE 辅助工具),由 AI 驱动,将自然语言需求自动转化为完整、CI/CD就绪的固件工程,辅助 Automotive SPICE SWE.1~SWE.6 合规证据准备。它用 AI 辅助流水线替代了需求工程、代码生成、审查、测试规划和合规证据收集中繁琐的人工环节。

一句话: yuleOSH 接收需求描述,输出经过审查、测试、CI集成的固件,并附带 ASPICE 合规辅助证据——AI 辅助完成。


快速开始

# 第一步:安装(15秒)
pip install yuleosh

# 第二步:初始化项目(15秒)
yuleosh init my-project

# 第三步:运行完整流水线(90秒)
cd my-project && yuleosh pipeline run docs/spec.md

三行命令,两分钟,从零到固件。


核心特性

🧠 OpenSpec 规范引擎

结构化需求,使用 RFC 2119 关键字(SHALL/SHOULD/MAY)配合 GIVEN/WHEN/THEN 场景。自动验证、差异对比、全链路追溯。

🔍 AI 代码审查

四代理并行审查矩阵覆盖架构、领域正确性、代码风格和测试覆盖率。8项嵌入式C静态分析 + 资源使用预测。

🔧 硬件在环

内置 OpenOCD(STM32)、JLink(ARM Cortex-M)、esptool(ESP32)适配器。一条命令即可自动刷写、监视串口、启动GDB调试。

☁️ SaaS 管理面板

Next.js 管理面板 + PostgreSQL 多租户存储 + JWT 认证 + 组织/项目隔离 + 流水线实时监控。

📋 合规审计

一键生成追溯矩阵、验收矩阵和合规证据 ZIP 包——辅助 ASPICE SWE.1~SWE.6 证据准备。

AI 辅助流水线

用户需求 → OpenSpec → 系统设计 → 详细设计 → 代码生成 → 内审 →
测试规划 → 代码审查 → CI运行 → 证据打包 → 部署

架构

[用户需求] ──▶ [OpenSpec 引擎] ──▶ [代理流水线] ──▶ [代码生成]
                                                    │
                                                    ▼
                                          验证 → CI → 硬件部署

四层架构细节参见英文版上方说明。


支持平台

平台 刷写工具 调试器
ESP32 / ESP32-S3 esptool idf-monitor + GDB
STM32 (F4/H7/G0) OpenOCD OpenOCD + GDB
ARM Cortex-M 系列 JLinkExe JLinkGDBServer
自定义平台 插件 API 插件 API

目录结构

yuleOSH/
├── src/yuleosh/    核心源码模块
├── frontend/       Next.js SaaS 管理面板
├── tests/          250+ 测试(全部通过)
├── docs/           需求文档、指南、报告
├── deploy/         生产部署配置
├── Dockerfile      多阶段 Docker 构建
├── docker-compose.yml  生产 Docker Compose
├── install.sh      一键安装脚本
└── pyproject.toml  Python 包配置

生产部署

Docker Compose(推荐)

git clone https://github.com/frisky1985/yuleOSH.git
cd yuleOSH
cp deploy/.env.production.example deploy/.env.production
# 编辑 deploy/.env.production 填入密钥
docker compose -f deploy/docker-compose.yml up -d

pip 安装(CLI 模式)

pip install yuleosh
yuleosh init my-project
yuleosh pipeline run docs/spec.md

一键安装(完整套件)

curl -fsSL https://raw.githubusercontent.com/frisky1985/yuleOSH/main/install.sh | bash

源码安装

git clone https://github.com/frisky1985/yuleOSH.git
cd yuleOSH
pip install -e .
yuleosh init .
yuleosh help

定价与版本

yuleOSH 提供三个版本。完整功能对比详见 版本分界线 · 功能矩阵 →

版本 定价 适用场景
社区版 (MIT) ¥0 个人开发者、开源项目
SaaS Pro ¥599/月 (¥5,999/年) 嵌入式合规团队,全功能流水线
企业版 ¥98,000/年 大型企业,私有化部署 + 人民币合同支持

📖 完整版本矩阵 →


路线图

版本 重点 状态
v0.1.0 基础—OpenSpec、代理流水线、CI/CD、证据
v0.2.0 ASPICE合规—严格模式、双向追溯
v0.3.0 地基加固—测试规划、层级、交叉编译
v1.0.0 生产就绪—HIL适配器、插件市场、扩展 🚧
v1.1.0 企业版—RBAC、审计日志、SAML SSO 📋
v1.2.0 云端—多区域、数据驻留、托管服务 📋

参与贡献

参见 CONTRIBUTING.md 了解开发环境配置、代码规范和 PR 流程。

安全

参见 SECURITY.md 了解漏洞披露流程。


法律免责声明

yuleOSH 是一款 ASPICE 合规辅助工具,不替代正式的 ASPICE 认证评估。合规状态取决于组织级项目管理流程,工具生成的结果仅供参考。


许可证

MIT 许可证 — 详见 LICENSE。Copyright (c) 2025 frisky1985。


为认真交付优质固件的嵌入式团队而构建。

About

yuleOSH - 嵌入式AI开发全流程平台:OpenSpec+Superpowers+Harness Engineering 三位一体,一站式需求→开发→测试持续集成/测试,AI Agents流水线

Resources

Code of conduct

Contributing

Security policy

Stars

Watchers

Forks

Releases

Packages

Used by

Contributors

Languages