{"id":41367206,"url":"https://github.com/heguangyong/kiro-spec-engine","last_synced_at":"2026-02-15T17:31:21.898Z","repository":{"id":334160156,"uuid":"1140273307","full_name":"heguangyong/kiro-spec-engine","owner":"heguangyong","description":"Kiro Spec Engine - A spec-driven development engine with steering rules and quality enhancement powered by Ultrawork spirit","archived":false,"fork":false,"pushed_at":"2026-02-12T18:26:20.000Z","size":3760,"stargazers_count":20,"open_issues_count":0,"forks_count":1,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-02-13T00:09:54.015Z","etag":null,"topics":["cli","development-tools","kiro","quality-enhancement","spec","ultrawork"],"latest_commit_sha":null,"homepage":"","language":"JavaScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/heguangyong.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-01-23T03:46:36.000Z","updated_at":"2026-02-12T18:37:14.000Z","dependencies_parsed_at":"2026-02-10T04:04:08.375Z","dependency_job_id":null,"html_url":"https://github.com/heguangyong/kiro-spec-engine","commit_stats":null,"previous_names":["heguangyong/kiro-spec-engine"],"tags_count":94,"template":false,"template_full_name":null,"purl":"pkg:github/heguangyong/kiro-spec-engine","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/heguangyong%2Fkiro-spec-engine","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/heguangyong%2Fkiro-spec-engine/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/heguangyong%2Fkiro-spec-engine/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/heguangyong%2Fkiro-spec-engine/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/heguangyong","download_url":"https://codeload.github.com/heguangyong/kiro-spec-engine/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/heguangyong%2Fkiro-spec-engine/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29414277,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-13T06:24:03.484Z","status":"ssl_error","status_checked_at":"2026-02-13T06:23:12.830Z","response_time":78,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.5:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["cli","development-tools","kiro","quality-enhancement","spec","ultrawork"],"created_at":"2026-01-23T08:35:08.743Z","updated_at":"2026-02-15T17:31:21.890Z","avatar_url":"https://github.com/heguangyong.png","language":"JavaScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# kse - Kiro Spec Engine\n\n[![npm version](https://badge.fury.io/js/kiro-spec-engine.svg)](https://badge.fury.io/js/kiro-spec-engine)\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)\n\n\u003e **⚠️ Important Clarification**: `kiro-spec-engine` (kse) is an **npm package and CLI tool** for spec-driven development.  \n\u003e It is **NOT** the Kiro IDE desktop application. If you're looking for Kiro IDE, visit https://kiro.dev\n\n**A context provider for AI coding tools** - Structure your project requirements, design, and tasks so AI assistants can help you build better software.\n\n**🚀 NEW: Autonomous Control** - Let AI independently manage entire development workflows from requirements to delivery.\n\nEnglish | [简体中文](README.zh.md)\n\n---\n\n## What is kse?\n\n**kse (Kiro Spec Engine) is a context management system for AI-assisted development.** It helps you organize project information into structured \"Specs\" (Requirements → Design → Tasks) that AI tools can understand and use effectively.\n\nThink of kse as a **librarian for your AI assistant** - it organizes and presents project context so your AI tool knows exactly what you're building, why, and how.\n\n### How it Works\n\n```mermaid\ngraph LR\n    A[You create Specs] --\u003e B[kse organizes context]\n    B --\u003e C[AI tool reads context]\n    C --\u003e D[AI generates better code]\n```\n\n1. **You create Specs** - Write requirements, design, and tasks in structured markdown files\n2. **kse organizes context** - Exports formatted context optimized for AI tools\n3. **AI tool reads context** - Your AI assistant understands your project structure\n4. **AI generates better code** - Code that matches your design and requirements\n\n### What kse is NOT\n\n- ❌ **Not a code generator** - kse doesn't write code; your AI tool does\n- ❌ **Not an IDE** - kse works alongside your existing development tools\n- ❌ **Not AI-specific** - Works with Claude, Cursor, Windsurf, Copilot, and any AI tool\n- ❌ **Not a replacement for documentation** - It's a structured way to organize project context\n\n### Who Should Use kse?\n\n- ✅ Developers using AI coding assistants (Claude, Cursor, Copilot, etc.)\n- ✅ Teams wanting structured project documentation\n- ✅ Anyone building features that need clear requirements and design\n- ✅ Projects that benefit from spec-driven development\n\n## Why kse Wins in AI Delivery\n\n| Advantage | kse Capability | Practical Impact |\n| --- | --- | --- |\n| Structure-first execution | Requirements → Design → Tasks + gate checks | Lower rework and fewer requirement drifts |\n| Autonomous close-loop delivery | One-command close loop (`auto close-loop`) | No step-by-step confirmation waits |\n| Multi-agent scale-out | DAG orchestration (`orchestrate run/status/stop`) | Parallel delivery without manual terminal juggling |\n| Auto portfolio decomposition | Master/Sub spec split + dependency wiring | Complex goals become executable parallel plans automatically |\n| Program self-healing closure | `auto close-loop-program` auto recovery + learned remediation memory | Failed/partial runs continue autonomously until bounded convergence |\n| Queue-driven autonomous runtime | `auto close-loop-controller` persistent queue drain | Program-scale backlogs continue without manual re-triggering |\n| Autonomous convergence governance | Program gate (`min-success-rate` + `max-risk-level`) + audit JSON output | Program completion is policy-verified and traceable |\n| Measurable outcomes | KPI automation (`value metrics snapshot/baseline/trend`) | Delivery quality can be tracked week-over-week |\n| Tool-agnostic adoption | Works across Claude/Cursor/Windsurf/Copilot/Kiro | No lock-in to a single AI IDE |\n| Built-in governance | Docs governance, lock management, audit, env/workspace controls | Team collaboration stays auditable and stable |\n\n### 90-Second Value Proof\n\n```bash\n# 1) Adopt kse in current repo\nkse adopt\n\n# 2) Generate Spec workflow draft\nkse spec bootstrap --name 01-00-demo-feature --non-interactive\n\n# 3) Generate KPI input sample\nkse value metrics sample --out ./kpi-input.json --json\n\n# 4) Produce machine-readable KPI snapshot\nkse value metrics snapshot --input ./kpi-input.json --json\n```\n\n---\n\n## Quick Start\n\n### The Simplest Way (30 seconds) ⚡\n\n**Just tell your AI:**\n\n```\nInstall kse and use it to manage this project with Spec-driven development.\n```\n\n**Your AI will:**\n1. Install kse globally (`npm install -g kiro-spec-engine`)\n2. Adopt it in your project (`kse adopt`)\n3. Read the methodology guide (`.kiro/README.md`)\n4. Start working according to Spec-driven approach\n\n**That's it!** Your AI handles everything. No manual steps needed.\n\n---\n\n### Step-by-Step Guide (if you want details) 📋\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cb\u003eClick to expand detailed steps\u003c/b\u003e\u003c/summary\u003e\n\n#### Step 1: Install kse (30 seconds)\n\n```bash\nnpm install -g kiro-spec-engine\n```\n\nVerify installation:\n```bash\nkse --version\n```\n\n#### Step 2: Adopt kse in Your Project (30 seconds)\n\nNavigate to your project directory and run:\n\n```bash\ncd your-project\nkse adopt\n```\n\nThis creates a `.kiro/` directory with:\n- `README.md` - Project development guide for AI\n- `specs/` - Where your Specs live\n- `steering/` - Development rules (optional)\n\n#### Step 3: Tell Your AI About the Methodology (30 seconds)\n\n**In your AI tool (Cursor, Claude, Windsurf, Kiro, etc.), say:**\n\n```\nPlease read .kiro/README.md to understand how this project works.\n```\n\n**Your AI will learn:**\n- This project follows Spec-driven development\n- Every feature starts with a Spec (requirements + design + tasks)\n- How to work with this methodology\n- When to use kse commands\n\n#### Step 4: Start Building Features\n\n**Just ask your AI to implement features naturally:**\n\n```\nI need a user login feature with email and password.\n```\n\n**Your AI will automatically:**\n1. Create a Spec with requirements, design, and tasks\n2. Implement according to the Spec\n3. Update task status as work progresses\n4. Use kse commands internally (you don't need to run them)\n\n**Example conversation:**\n- **You**: \"I need user login with email and password\"\n- **AI**: \"I'll create a Spec for this. Let me define the requirements...\"\n- **AI**: \"Here's the design... Now I'll implement task 1.1...\"\n- **AI**: \"Task 1.1 complete. Moving to task 1.2...\"\n\n\u003c/details\u003e\n\n---\n\n**Key insight:** You don't \"use kse\" - your project \"follows Spec-driven methodology\" and kse helps enforce it. The AI handles all the kse commands for you.\n\n### Step 5: Next Steps (30 seconds)\n\n- 📖 Read the [Quick Start Guide](docs/quick-start.md) for detailed examples\n- 🔧 Check your tool's integration guide: [Cursor](docs/tools/cursor-guide.md) | [Claude](docs/tools/claude-guide.md) | [Windsurf](docs/tools/windsurf-guide.md)\n- 💡 Learn about [Integration Modes](docs/integration-modes.md)\n\n---\n\n## Core Concepts\n\n### Specs\n\nA **Spec** is a structured description of a feature or project component. Each Spec contains:\n\n- **Requirements** (`requirements.md`) - What you're building and why\n- **Design** (`design.md`) - How you'll build it (architecture, APIs, components)\n- **Tasks** (`tasks.md`) - Step-by-step implementation checklist\n\n### Context Export\n\n**Context export** transforms your Spec into a format optimized for AI tools. It includes:\n- All requirements, design decisions, and tasks\n- Project structure and conventions\n- Steering rules (optional) for AI behavior\n\n### Integration Modes\n\nkse supports three ways to work with AI tools:\n\n1. **Native Integration** - AI tool directly accesses kse (Kiro IDE)\n2. **Manual Export** - You export and paste context (Claude, ChatGPT, Cursor)\n3. **Watch Mode** - Automatic context updates on file changes (all tools)\n\nLearn more: [Integration Modes Guide](docs/integration-modes.md)\n\n---\n\n## Integration with Your AI Tool\n\nkse works with any AI coding assistant. Choose your tool for specific guidance:\n\n### Popular AI Tools\n\n- **[Cursor](docs/tools/cursor-guide.md)** - IDE with AI pair programming\n- **[Claude Code](docs/tools/claude-guide.md)** - Anthropic's coding assistant\n- **[Windsurf](docs/tools/windsurf-guide.md)** - AI agent with command execution\n- **[Kiro](docs/tools/kiro-guide.md)** - Native integration, no manual export needed\n- **[VS Code + Copilot](docs/tools/vscode-guide.md)** - GitHub Copilot integration\n- **[Generic AI Tools](docs/tools/generic-guide.md)** - Works with any AI assistant\n\n### Integration Workflow\n\n```mermaid\nsequenceDiagram\n    participant You\n    participant AI Tool\n    participant kse\n    You-\u003e\u003eAI Tool: \"I have a Spec for user-login, implement task 1.1\"\n    AI Tool-\u003e\u003ekse: kse context export user-login\n    kse-\u003e\u003eAI Tool: Spec content (requirements, design, tasks)\n    AI Tool-\u003e\u003eAI Tool: Generate code following Spec\n    AI Tool-\u003e\u003eYou: Here's the implementation\n    AI Tool-\u003e\u003ekse: Update tasks.md (mark task complete)\n```\n\n**Key insight:** You stay in your AI tool. The AI reads the Spec and generates code that matches your design.\n\n---\n\n## Documentation\n\n### Getting Started\n- 📖 **[Quick Start Guide](docs/quick-start.md)** - Detailed 5-minute tutorial\n- 🤔 **[FAQ](docs/faq.md)** - Frequently asked questions\n- 🔧 **[Troubleshooting](docs/troubleshooting.md)** - Common issues and solutions\n\n### Core Guides\n- 📋 **[Spec Workflow](docs/spec-workflow.md)** - Understanding Specs in depth\n- 🔢 **[Spec Numbering Strategy](docs/spec-numbering-guide.md)** - How to number your Specs\n- 📄 **[Document Governance](docs/document-governance.md)** - Automated document management\n- 🌍 **[Environment Management](docs/environment-management-guide.md)** - Multi-environment configuration\n- 📦 **[Multi-Repository Management](docs/multi-repo-management-guide.md)** - Manage multiple Git repositories\n- 🎭 **[Scene Runtime](docs/scene-runtime-guide.md)** - Template engine, quality pipeline, ontology, Moqui ERP\n- 🤖 **[Multi-Agent Coordination](docs/multi-agent-coordination-guide.md)** - Parallel agent coordination\n- 📈 **[Value Observability](docs/value-observability-guide.md)** - KPI snapshot, baseline, trend, gate-ready evidence\n- 🔌 **[Integration Modes](docs/integration-modes.md)** - Three ways to integrate kse\n- 📝 **[Command Reference](docs/command-reference.md)** - All kse commands\n\n### Tool-Specific Guides\n- [Cursor Integration](docs/tools/cursor-guide.md)\n- [Claude Code Integration](docs/tools/claude-guide.md)\n- [Windsurf Integration](docs/tools/windsurf-guide.md)\n- [Kiro Integration](docs/tools/kiro-guide.md)\n- [VS Code + Copilot Integration](docs/tools/vscode-guide.md)\n- [Generic AI Tools](docs/tools/generic-guide.md)\n\n### Examples\n- [API Feature Example](docs/examples/add-rest-api/) - RESTful API Spec\n- [UI Feature Example](docs/examples/add-user-dashboard/) - React dashboard Spec\n- [CLI Feature Example](docs/examples/add-export-command/) - CLI command Spec\n\n### Advanced Topics\n- [Adoption Guide](docs/adoption-guide.md) - Adopting kse in existing projects\n- [Upgrade Guide](docs/upgrade-guide.md) - Version upgrade instructions\n- [Release Archive](docs/releases/README.md) - Versioned release notes and validation report index\n- [Release Notes v1.46.2](docs/releases/v1.46.2.md) - Current release highlights and impact\n- [Validation Report v1.46.2](docs/releases/v1.46.2-validation.md) - Release-readiness evidence and verification results\n- [Release Checklist](docs/release-checklist.md) - Repeatable pre-release verification flow\n- [Manual Workflows](docs/manual-workflows-guide.md) - Step-by-step workflows\n- [Developer Guide](docs/developer-guide.md) - Contributing and extending kse\n\n### Complete Documentation\n- 📚 **[Documentation Index](docs/README.md)** - All documentation in one place\n\n---\n\n## Key Features\n\n### Autonomous Control 🚀 NEW in v1.23.0\n- **Fully Autonomous Execution**: AI independently manages entire development workflows from requirements to delivery\n- **Closed-Loop Program Execution**: `kse auto close-loop \"\u003cgoal\u003e\"` runs from goal decomposition to terminal orchestration state without manual confirmation gates\n- **Automatic Master/Sub Portfolio Split**: Large goals are auto-split into dependency-wired sub-specs and coordinated by a master spec\n- **Semantic Decomposition + Live Stream**: Clause/category-aware goal parsing plus live orchestration status stream (disable via `--no-stream`)\n- **Session Resume + Retention Hygiene**: Resume interrupted runs with `--resume latest` and maintain archives via `kse auto session list/prune`, `kse auto batch-session list/prune`, and `kse auto controller-session list/prune`\n- **Intelligent Error Recovery**: Automatically diagnose and fix errors with 3 retry attempts and learning system\n- **Strategic Checkpoints**: Pause only at meaningful milestones (phase boundaries, fatal errors, external resources)\n- **Continuous Task Execution**: Execute multiple tasks without interruption between individual tasks\n- **Learning System**: Improve over time by learning from successful and failed recovery strategies\n- **Safety Boundaries**: Respect workspace boundaries, require confirmation for production/external operations\n- **Three Execution Modes**: Conservative (safe), Balanced (default), Aggressive (fast)\n- **Progress Tracking**: Real-time status, estimated completion time, detailed execution reports\n- **Rollback Support**: Create checkpoints and rollback to previous states if needed\n- **CORE_PRINCIPLES Compliance**: Follows Spec-driven development, file management, and quality standards\n\n**Quick Start**:\n```bash\n# Decompose one broad goal into master/sub specs and run to closure\nkse auto close-loop \"build autonomous close-loop and master/sub multi-spec execution for kse\"\n\n# Create and execute a feature autonomously\nkse auto create \"user authentication with JWT tokens\"\n\n# Run existing Spec autonomously\nkse auto run 33-00-ai-autonomous-control\n\n# Check status\nkse auto status\n\n# Resume after pause\nkse auto resume\n```\n\n[Learn more about Autonomous Control →](docs/autonomous-control-guide.md)\n\n### Spec-Driven Development\nStructure your work with Requirements → Design → Tasks workflow\n\n### KPI Automation \u0026 Observability 🚀 NEW in v1.46.2\n- **Unified Metric Contract**: Load and validate KPI definitions from `metric-definition.yaml`\n- **Weekly Snapshot Pipeline**: Generate machine-readable snapshots with risk level and audit reasons\n- **Baseline and Trend Analysis**: Build baseline from historical data and detect worsening trends automatically\n- **Gate-Ready Summary**: Emit Day30/Day60-consumable summary payloads with evidence paths\n- **CLI Commands**: `value metrics sample`, `value metrics snapshot`, `value metrics baseline`, `value metrics trend` with `--json` support\n\n### Multi-Workspace Management 🚀 NEW in v1.11.0\n- **Workspace Registry**: Manage multiple kse projects from a single location\n- **Quick Switching**: Switch between projects without directory navigation\n- **Data Atomicity**: Single source of truth (`~/.kse/workspace-state.json`)\n- **Cross-Platform**: Consistent path handling across Windows/Linux/macOS\n- **Auto Migration**: Seamless upgrade from legacy workspace format\n\n### Environment Configuration Management 🚀 NEW in v1.14.0\n- **Environment Registry**: Manage multiple environment configurations (dev, test, staging, prod)\n- **Quick Switching**: Switch between environments with automatic backup\n- **Automatic Backup**: Create timestamped backups before each switch\n- **Rollback Support**: Restore previous environment configuration instantly\n- **Verification**: Validate environment configuration after switching\n- **Command Execution**: Run commands in specific environment context\n- **Cross-Platform**: Works seamlessly on Windows, Linux, and macOS\n\n### Multi-Repository Management 🚀 NEW in v1.20.0\n- **Unified Interface**: Manage multiple Git subrepositories from a single command\n- **Auto-Discovery**: Automatically scan and configure all Git repositories in your project\n- **Nested Repository Support**: Discover and manage Git repositories nested inside other repositories\n- **Parent-Child Tracking**: Track relationships between parent and nested repositories\n- **Batch Operations**: Execute Git commands across all repositories simultaneously\n- **Status Overview**: View status of all repositories in a single table with parent relationships\n- **Health Checks**: Verify repository configuration and connectivity\n- **Cross-Platform**: Consistent path handling across Windows/Linux/macOS\n- **Smart Exclusions**: Automatically skip common non-repository directories (node_modules, build, etc.)\n\n### Moqui ERP Integration 🚀 NEW in v1.39.0\n- **Moqui ERP Adapter**: Connect KSE scene runtime to live Moqui ERP instances\n  - `MoquiClient` — HTTP client with JWT auth lifecycle (login, refresh, re-login, logout) and retry logic\n  - `MoquiAdapter` — Binding handler for `spec.erp.*` and `moqui.*` refs, entity CRUD, service invocation, screen discovery\n- **Scene Template Extractor** (v1.40.0): Analyze Moqui resources, identify business patterns, generate reusable scene templates\n  - Entity grouping by Header/Item suffix patterns (e.g., OrderHeader + OrderItem → composite)\n  - Pattern-based manifest generation with governance contracts\n- **CLI Commands**: `scene connect`, `scene discover`, `scene extract` with `--json` support\n\n### Multi-Agent Parallel Coordination 🚀 NEW in v1.43.0\n- **Agent Registry**: MachineIdentifier-based agent lifecycle with heartbeat monitoring and inactive cleanup\n- **Task Lock Manager**: File-based task locking with atomic operations, single-agent backward compatibility\n- **Task Status Store**: Concurrent-safe task status updates with exponential backoff retry\n- **Steering File Lock**: Write serialization with pending-file degradation fallback\n- **Merge Coordinator**: Git branch isolation per agent (`agent/{agentId}/{specName}`), conflict detection, auto-merge\n- **Central Coordinator**: Dependency-driven ready task computation, task assignment, progress tracking\n- **Zero Overhead**: All components are no-ops in single-agent mode (full backward compatibility)\n\n[Learn more about Multi-Agent Coordination →](docs/multi-agent-coordination-guide.md)\n\n### Agent Orchestrator 🚀 NEW in v1.45.0\n- **Automated Multi-Agent Spec Execution**: Replace manual multi-terminal workflows with a single command\n- **DAG-Based Scheduling**: Analyze Spec dependencies, compute execution batches via topological sort\n- **Parallel Execution**: Run up to N Specs simultaneously via Codex CLI sub-agents (`--max-parallel`)\n- **Failure Propagation**: Failed Spec's dependents automatically marked as skipped\n- **Retry Mechanism**: Configurable automatic retry for failed Specs\n- **Real-Time Monitoring**: Track per-Spec status and overall orchestration progress\n- **Graceful Termination**: Stop all sub-agents cleanly (SIGTERM → SIGKILL)\n- **Configurable**: Codex command, args, parallelism, timeout, retries via `.kiro/config/orchestrator.json`\n\n**Quick Start**:\n```bash\n# Run 3 Specs in parallel via Codex CLI\nkse orchestrate run --specs \"spec-a,spec-b,spec-c\" --max-parallel 3\n\n# Check orchestration progress\nkse orchestrate status\n\n# Stop all sub-agents\nkse orchestrate stop\n```\n\nTip: `kse spec bootstrap|pipeline run|gate run --specs ...` now defaults to this orchestrate mode automatically.\n\n**Recommended Codex-Orchestrator config (`.kiro/config/orchestrator.json`)**:\n```json\n{\n  \"agentBackend\": \"codex\",\n  \"maxParallel\": 3,\n  \"timeoutSeconds\": 900,\n  \"maxRetries\": 2,\n  \"apiKeyEnvVar\": \"CODEX_API_KEY\",\n  \"codexArgs\": [\"--skip-git-repo-check\"],\n  \"codexCommand\": \"npx @openai/codex\"\n}\n```\n\nIf Codex CLI is globally installed, you can set `\"codexCommand\": \"codex\"`.\n\n### Spec-Level Steering \u0026 Context Sync 🚀 NEW in v1.44.0\n- **Spec Steering (L4)**: Independent `steering.md` per Spec with constraints, notes, and decisions — zero cross-agent conflict\n- **Steering Loader**: Unified L1-L4 four-layer steering loader with priority-based merging\n- **Context Sync Manager**: Multi-agent friendly CURRENT_CONTEXT.md with structured Spec progress table, concurrent-safe updates\n- **Spec Lifecycle Manager**: State machine (planned → assigned → in-progress → completed → released) with auto-completion detection\n- **Sync Barrier**: Agent Spec-switch synchronization — uncommitted change check + steering reload\n- **Coordinator Integration**: Auto Spec completion check on task complete, SyncBarrier on task assign\n\n### Scene Ontology Enhancement 🚀 NEW in v1.42.0\n- **OntologyGraph**: Semantic relationship graph for binding refs (depends_on, composes, extends, produces)\n- **Action Abstraction**: Intent, preconditions, postconditions per binding for AI agent understanding\n- **Data Lineage**: Source → transform → sink tracking in governance_contract\n- **Agent-Ready Metadata**: `agent_hints` field with summary, complexity, duration, permissions, sequence\n- **Agent Readiness Score**: Bonus quality dimension (max +10 points)\n- **CLI Commands**: `scene ontology show|deps|validate|actions|lineage|agent-info` with `--json` support\n\n### Scene Template Quality Pipeline 🚀 NEW in v1.41.0\n- **Template Lint Engine**: 10-category quality checks for scene packages (manifest completeness, binding refs, governance, consistency, variables, documentation, action abstraction, data lineage, agent hints)\n- **Quality Score Calculator**: 4-dimension scoring + agent_readiness bonus (contract validity, lint pass rate, documentation quality, governance completeness + agent readiness max +10) with 0-100+ scale\n- **One-Stop Contribute Pipeline**: Validate → Lint → Score → Preview → Publish in a single command\n- **CLI Commands**: `scene lint`, `scene score`, `scene contribute` with `--strict`, `--dry-run`, `--skip-lint`, `--json` support\n\n### Scene Template Engine 🚀 NEW in v1.25.0\n- **Template Variable Schema**: Define typed variables (string, number, boolean, enum, array) with validation rules in scene-package.json\n- **Multi-File Rendering**: Recursive template processing with `{{variable}}` substitution, `{{#if}}` conditionals, `{{#each}}` loops\n- **Three-Layer Inheritance**: L1-Capability / L2-Domain / L3-Instance package hierarchy with schema and file merging\n- **CLI Commands**: `scene template-validate`, `scene template-resolve`, `scene template-render` with `--json` support\n\n### Spec-Level Collaboration 🚀 NEW in v1.22.0\n- **Parallel Development**: Enable multiple AI instances to work on different Specs simultaneously\n- **Dependency Management**: Define and track dependencies between Specs with circular dependency detection\n- **Interface Contracts**: Formal API definitions ensuring compatibility between independently developed Specs\n- **Status Tracking**: Monitor progress and assignments across all Specs in real-time\n- **Integration Testing**: Run cross-Spec integration tests to verify modules work together\n- **Dependency Visualization**: View dependency graphs with critical path highlighting\n- **Backward Compatible**: Opt-in system that doesn't affect existing single-Spec workflows\n\n[Learn more about Spec-Level Collaboration →](docs/spec-collaboration-guide.md)\n\n### Spec Locking Mechanism 🚀 NEW\n- **Multi-User Coordination**: Prevent conflicts when multiple developers work on the same Spec\n- **Machine Identification**: Unique machine IDs for accurate lock ownership tracking\n- **Stale Lock Detection**: Automatic detection and cleanup of abandoned locks (default: 24h timeout)\n- **Lock Status Integration**: View lock status in `kse status` output\n- **Force Unlock**: Override locks when necessary with `--force` flag\n- **Backward Compatible**: Locking is optional and doesn't affect existing workflows\n\n**Quick Start**:\n```bash\n# Acquire a lock before editing\nkse lock acquire my-feature --reason \"Implementing auth module\"\n\n# Check lock status\nkse lock status\n\n# Release when done\nkse unlock my-feature\n\n# Clean up stale locks\nkse lock cleanup\n\n# View your machine ID\nkse lock whoami\n```\n\n### DevOps Integration Foundation 🚀\n- **Operations Spec Management**: Standardized operations documentation (deployment, monitoring, troubleshooting, etc.)\n- **Progressive AI Autonomy**: L1-L5 takeover levels for gradual AI operations control\n- **Audit Logging**: Tamper-evident audit trail with SHA-256 integrity verification\n- **Feedback Integration**: Automated user feedback processing and analytics\n- **Permission Management**: Environment-based security controls (dev, test, pre-prod, prod)\n- **Operations Validation**: Complete spec validation with clear error reporting\n\n### Document Governance\n- Automated document lifecycle management\n- Clean project structure enforcement\n- Temporary file cleanup\n- Artifact organization\n- Git hooks for compliance\n\n### Multi-User Collaboration\n- Personal workspaces for team members\n- Task claiming and tracking\n- Workspace synchronization\n\n### Cross-Tool Compatibility\nExport context for Claude Code, Cursor, Windsurf, Copilot, and more\n\n### Watch Mode Automation\nAutomatic file monitoring and context updates\n\n### Quality Enhancement\n- Document quality scoring (0-10 scale)\n- Intelligent improvement suggestions\n- Professional standards enforcement\n\n### Multi-Language Support\nEnglish and Chinese interfaces\n\n---\n\n## Command Overview\n\n```bash\n# Project setup\nkse adopt                          # Adopt kse in existing project\nkse create-spec \u003cname\u003e             # Legacy: create empty Spec folder only\n\n# Autonomous close-loop program (one command, no confirmation loop)\nkse auto close-loop \"\u003cgoal\u003e\"       # Auto split goal into master/sub specs and execute to closure\nkse auto close-loop \"\u003cgoal\u003e\" --dry-run --json  # Preview decomposition plan only\nkse auto close-loop-program \"\u003cgoal\u003e\" --program-govern-until-stable --program-govern-use-action 1 --json # Program-level recovery + governance with remediation action execution\nkse auto close-loop-controller .kiro/auto/program-queue.lines --wait-on-empty --dequeue-limit 2 --json # Queue-driven autonomous controller for broad-goal backlogs\nkse auto close-loop-controller --controller-resume latest --json # Resume autonomous controller from latest persisted checkpoint\nkse auto controller-session list --limit 50 --json # Inspect persisted close-loop-controller summary sessions\nkse auto controller-session prune --keep 20 --older-than-days 14 --dry-run --json # Prune old controller summaries by retention policy\n\n# Spec workflow (recommended)\nkse spec bootstrap --name \u003cspec\u003e --non-interactive         # Generate requirements/design/tasks draft\nkse spec pipeline run --spec \u003cspec\u003e                         # Run staged workflow for one Spec\nkse spec gate run --spec \u003cspec\u003e --json                      # Run standardized Spec gate checks\nkse spec bootstrap --specs \"\u003cspec-a,spec-b\u003e\" --max-parallel \u003cN\u003e  # Multi-Spec defaults to orchestrate\nkse spec pipeline run --specs \"\u003cspec-a,spec-b\u003e\" --max-parallel \u003cN\u003e # Multi-Spec defaults to orchestrate\nkse spec gate run --specs \"\u003cspec-a,spec-b\u003e\" --max-parallel \u003cN\u003e     # Multi-Spec defaults to orchestrate\n\n# Context management\nkse context export \u003cspec-name\u003e     # Export context for AI tools\nkse prompt generate \u003cspec\u003e \u003ctask\u003e  # Generate task-specific prompt\n\n# KPI automation and observability (NEW in v1.46.2)\nkse value metrics sample --out \u003cpath\u003e --json               # Generate sample KPI input JSON\nkse value metrics snapshot --input \u003cpath\u003e --json          # Generate weekly KPI snapshot + gate summary\nkse value metrics baseline --from-history \u003cN\u003e --json      # Build baseline from earliest N snapshots\nkse value metrics trend --window \u003cN\u003e --json               # Analyze trend/risk from latest N snapshots\nkse auto kpi trend --weeks 8 --period week --json         # Aggregate autonomous delivery KPI trend (weekly/daily buckets + anomalies)\nkse auto kpi trend --weeks 8 --period day --csv --out \u003cpath\u003e # Export autonomous KPI trend as CSV\nkse auto kpi trend --weeks 8 --mode controller --json     # Filter trend to close-loop-controller sessions only\n\n# Workspace management (NEW in v1.11.0)\nkse workspace create \u003cname\u003e [path] # Register a new workspace\nkse workspace list                 # List all workspaces\nkse workspace switch \u003cname\u003e        # Switch active workspace\nkse workspace info [name]          # Show workspace details\nkse workspace remove \u003cname\u003e        # Remove workspace\n\n# Environment management (NEW in v1.14.0)\nkse env list                       # List all environments\nkse env switch \u003cname\u003e              # Switch to environment (with backup)\nkse env info                       # Show active environment details\nkse env register \u003cconfig-file\u003e     # Register new environment\nkse env unregister \u003cname\u003e          # Remove environment\nkse env rollback                   # Rollback to previous environment\nkse env verify                     # Verify current environment\nkse env run \"\u003ccommand\u003e\"            # Run command in environment context\n\n# Multi-repository management (NEW in v1.20.0)\nkse repo init [--nested]           # Initialize repository configuration (nested scanning by default)\nkse repo init --no-nested          # Initialize without nested repository scanning\nkse repo status [--verbose]        # Show status of all repositories (including nested)\nkse repo exec \"\u003ccommand\u003e\"          # Execute command in all repositories\n\n# Spec-level collaboration (NEW in v1.22.0)\nkse collab init \u003cmaster\u003e [options] # Initialize Master Spec with Sub-Specs\nkse collab status [spec] [--graph] # Display collaboration status\nkse collab assign \u003cspec\u003e \u003ckiro\u003e    # Assign Spec to Kiro instance\nkse collab verify \u003cspec\u003e           # Verify interface contracts\nkse collab integrate \u003cspecs...\u003e    # Run integration tests\nkse collab migrate \u003cspec\u003e          # Convert standalone Spec to collaborative\nkse repo health                    # Check repository health\n\n# Scene template engine (NEW in v1.25.0)\nkse scene template-validate --package \u003cpath\u003e   # Validate template variable schema\nkse scene template-resolve --package \u003cname\u003e    # Resolve inheritance chain and merged schema\nkse scene template-render --package \u003cname\u003e --values \u003cjson\u003e --out \u003cdir\u003e  # Render template files\n\n# Moqui ERP integration (NEW in v1.39.0)\nkse scene connect --config \u003cpath\u003e              # Test connectivity to Moqui ERP instance\nkse scene discover --config \u003cpath\u003e             # Discover entities, services, screens from Moqui\nkse scene extract --config \u003cpath\u003e --out \u003cdir\u003e  # Extract scene templates from Moqui (v1.40.0)\n\n# Scene template quality pipeline (NEW in v1.41.0)\nkse scene lint --package \u003cpath\u003e                # Lint scene package for quality issues\nkse scene score --package \u003cpath\u003e               # Calculate quality score (0-100)\nkse scene contribute --package \u003cpath\u003e          # One-stop validate → lint → score → publish\n\n# Scene ontology (NEW in v1.42.0)\nkse scene ontology show --package \u003cpath\u003e       # Show ontology graph\nkse scene ontology deps --ref \u003cref\u003e            # Query dependency chain\nkse scene ontology validate --package \u003cpath\u003e   # Validate graph consistency\nkse scene ontology actions --ref \u003cref\u003e         # Show action abstraction\nkse scene ontology lineage --ref \u003cref\u003e         # Show data lineage\nkse scene ontology agent-info --package \u003cpath\u003e # Show agent hints\nkse scene contribute --package \u003cpath\u003e          # One-stop validate → lint → score → publish\n\n# Agent orchestration (NEW in v1.45.0)\nkse orchestrate run --specs \"\u003cspec-list\u003e\" --max-parallel \u003cN\u003e  # Start multi-agent orchestration\nkse orchestrate status                         # View orchestration progress\nkse orchestrate stop                           # Stop all sub-agents\n\n# DevOps operations\nkse ops init \u003cproject-name\u003e        # Initialize operations specs\nkse ops validate [\u003cproject\u003e]       # Validate operations completeness\nkse ops audit [options]            # Query audit logs\nkse ops takeover \u003caction\u003e          # Manage AI takeover levels\nkse ops feedback \u003caction\u003e          # Manage user feedback\n\n# Task management\nkse task claim \u003cspec\u003e \u003ctask-id\u003e    # Claim a task\nkse task list \u003cspec\u003e               # List claimed tasks\n\n# Spec locking (multi-user collaboration)\nkse lock acquire \u003cspec-name\u003e       # Acquire lock on a Spec\nkse lock release \u003cspec-name\u003e       # Release lock (or use: kse unlock)\nkse lock status [spec-name]        # Show lock status\nkse lock cleanup                   # Remove stale locks\nkse lock whoami                    # Show machine identifier\nkse unlock \u003cspec-name\u003e [--force]   # Release lock (alias)\n\n# Document governance\nkse docs diagnose                  # Check document compliance\nkse docs cleanup                   # Remove temporary files\nkse docs validate                  # Validate document structure\nkse docs archive --spec \u003cname\u003e     # Organize Spec artifacts\nkse docs hooks install             # Install Git pre-commit hooks\n\n# Automation\nkse watch start                    # Start watch mode\nkse watch status                   # Check watch status\n\n# Project info\nkse status                         # Project status\nkse workflows                      # List available workflows\n```\n\nSee [Command Reference](docs/command-reference.md) for complete documentation.\n\n---\n\n## Contributing \u0026 Support\n\n### Getting Help\n\n- 📖 **Documentation**: Start with the [Quick Start Guide](docs/quick-start.md)\n- 🐛 **Bug Reports**: [GitHub Issues](https://github.com/heguangyong/kiro-spec-engine/issues)\n- 💬 **Discussions**: [GitHub Discussions](https://github.com/heguangyong/kiro-spec-engine/discussions)\n\n### Contributing\n\nWe welcome contributions! See our [Contributing Guide](CONTRIBUTING.md) for:\n- Code contributions\n- Documentation improvements\n- Bug reports and feature requests\n- Translation help\n\n### Development Setup\n\n```bash\ngit clone https://github.com/heguangyong/kiro-spec-engine.git\ncd kiro-spec-engine\nnpm install\nnpm link  # For local development\nnpm test  # Run tests\n```\n\n---\n\n## License\n\nMIT License - see [LICENSE](LICENSE) file for details.\n\n---\n\n## Acknowledgments\n\n- Inspired by the **Sisyphus** myth and the concept of noble struggle\n- Built on the foundation of **Kiro** spec-driven development\n- Influenced by **oh-my-opencode** and the Ultrawork Manifesto\n\n---\n\n**Ready to enhance your AI-assisted development?** 🚀\n\n```bash\nnpm install -g kiro-spec-engine\nkse adopt\nkse spec bootstrap --name 01-00-my-first-feature --non-interactive\n```\n\n---\n\n## 💬 Community \u0026 Discussion\n\nJoin our community to discuss AI-driven development, Spec workflows, and best practices.\n\n### WeChat Group\n\n\u003cimg src=\"docs/images/wechat-qr.png\" width=\"200\" alt=\"WeChat Group QR Code\"\u003e\n\n*Scan to add WeChat, note \"kse\" to join the group*\n\n### Other Channels\n\n- **GitHub Discussions**: [Join Discussion](https://github.com/heguangyong/kiro-spec-engine/discussions)\n- **Issues**: [Report Issues](https://github.com/heguangyong/kiro-spec-engine/issues)\n\n### Featured Article\n\n📖 **[The Philosophy and Practice of AI-Driven Development](docs/articles/ai-driven-development-philosophy-and-practice.md)**\n\nA deep conversation about AI development trends, Neo-Confucian philosophy, and software engineering practices. Based on 2000+ hours of AI-assisted programming experience.\n\n[Read in English](docs/articles/ai-driven-development-philosophy-and-practice.en.md) | [中文版](docs/articles/ai-driven-development-philosophy-and-practice.md)\n\n**Also available on:**\n- [WeChat Official Account (微信公众号)](https://mp.weixin.qq.com/s/GRo0XQ6GvQ03T4_FTvAsKA)\n- [Juejin (掘金)](https://juejin.cn/post/7598899986857377798)\n- [Zhihu (知乎)](https://zhuanlan.zhihu.com/p/1999164891391624163)\n- [X/Twitter](https://x.com/heguangyong/status/2015668235065229782)\n\n---\n\n**Version**: 1.47.4  \n**Last Updated**: 2026-02-15\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fheguangyong%2Fkiro-spec-engine","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fheguangyong%2Fkiro-spec-engine","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fheguangyong%2Fkiro-spec-engine/lists"}