菜单
Integrated project management system covering documentation, language initialization, template optimization, docs generation, and JIT document loading. Absorbed from moai-workflow-templates, moai-docs-generation, and moai-workflow-jit-docs.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
基于 SOC 职业分类
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/modu-ai/moai-adk --skill moai-workflow-project命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
Harness learning subsystem coordinator. Produces Tier 4 auto-update proposal payloads consumed by the orchestrator (which surfaces them via AskUserQuestion) and orchestrates Apply/Rollback flows. Triggers when harness learning proposals are pending or learning lifecycle management is needed.
DEPRECATED — legacy 7-Phase meta-harness. Redirects to the v4 harness Builder (/moai:harness <natural-language request>) which replaces the static 7-Phase workflow with an orchestrator-direct 4-phase Builder (ANALYZE / PLAN / GENERATE / ACTIVATE) + a manifest-driven dynamic-workflow Runner. Retained as the redirect source for backward-compat invocation paths; the 7-Phase body below is preserved as historical reference, NOT for new harness creation.
MoAI unified orchestrator for autonomous development. Routes natural language or subcommands (plan, run, sync, project, fix, loop, mx, feedback, review, clean, codemaps, gate, e2e, harness, goal, todo) to specialized agents.
| name | moai-workflow-project |
| description | Integrated project management system covering documentation, language initialization, template optimization, docs generation, and JIT document loading. Absorbed from moai-workflow-templates, moai-docs-generation, and moai-workflow-jit-docs. |
| when_to_use | Use for integrated project management: documentation scaffolding (product/structure/tech.md), multilingual language initialization, template and boilerplate optimization, docs generation (Sphinx, MkDocs, TypeDoc, OpenAPI), and JIT document loading. |
| license | Apache-2.0 |
| compatibility | Designed for Claude Code |
| allowed-tools | Read, Write, Edit, Bash(git:*), Bash(npm:*), Bash(npx:*), Bash(uv:*), Bash(pip:*), Bash(ls:*), Bash(mkdir:*), Grep, Glob, WebFetch, WebSearch |
| user-invocable | false |
| metadata |
| {"version":"3.0.0","category":"workflow","status":"active","updated":"2026-04-25","modularized":"true","tags":"workflow, project, documentation, initialization, templates, boilerplate, scaffolding, jit-docs, docs-generation","aliases":"moai-workflow-project","related-skills":"moai-workflow-spec, moai-workflow-docs-claim-check"} |
| progressive_disclosure | {"enabled":true,"level1_tokens":100,"level2_tokens":5000} |
Comprehensive project management system integrating documentation generation, multilingual support, and template optimization into a unified architecture with intelligent automation and Claude Code integration.
Scope: Consolidates documentation management, language initialization, and template optimization into a single cohesive system supporting complete project lifecycle from initialization to maintenance.
Target: Claude Code agents for project setup, documentation generation, multilingual support, and performance optimization.
Core Capabilities:
Key Features:
Supported Project Types: web applications, mobile applications, CLI tools, libraries, ML projects.
Three capability areas:
Three workflows: project initialization, documentation generation from SPEC, template performance optimization. Each follows a 3-step pattern (configure → execute → review results).
See core workflow walkthroughs for detailed step-by-step procedures.
Automatic Language Detection: analyzes file content, configuration files, system locale, and directory structure.
Multilingual Documentation: language-specific directories (e.g., docs/ko, docs/en), language negotiation, automatic redirection.
Agent Prompt Localization: language-specific instructions, cultural context, token cost optimization.
See language and localization detail for token cost analysis and locale configuration.
Performance Analysis: file size, complexity, performance bottlenecks, optimization opportunities, resource usage, backup recommendations.
Optimization Techniques: whitespace reduction, structure optimization, complexity reduction, performance caching.
Integrated configuration covers project metadata, language settings + costs, documentation status, template optimization results, module initialization states.
Language Settings: conversation_language (user-facing), agent_prompt_language (internal, often English for cost), documentation language (per language).
Updates trigger configuration file modifications, documentation structure updates, template localization.
See configuration schema and language fields for full field reference and supported language metadata.
For advanced patterns (custom templates, performance caching, batch processing, integration workflows), see references/reference.md and references/examples.md.
| Operation | Typical Duration |
|---|---|
| Complete documentation generation | 2-5 seconds |
| Language detection analysis | ~500 ms |
| Template optimization | 10-30 seconds |
| Configuration updates | ~100 ms |
Memory: base ~50MB, large projects +10-50MB, optimization cache 5-20MB.
File sizes: documentation 50-200KB per project, optimization backups match originals, configuration 5-10KB.
| Rationalization | Reality |
|---|---|
| "project docs are a one-time setup, no need to keep them current" | Stale product.md and tech.md mislead every SPEC written against them. They are living documents. |
| "structure.md will stay accurate since the codebase does not change much" | Every PR that adds a directory invalidates structure.md. Sync on every /moai project invocation. |
| "I know the tech stack, I do not need tech.md" | tech.md is not for you. It is for every agent that consults the project context before acting. |
| "codemaps take too long to generate" | codemaps are the only artifact that gives agents file-level awareness without reading every file. The cost pays for itself. |
| "I will write project docs after the feature is done" | Post-hoc docs capture what was built, not what was intended. Pre-feature docs guide the build. |
ls output of the project rootCode boilerplates, feedback templates, scaffolding, and project template optimization.
Analysis metrics: file size, complexity score, redundancy ratio, load performance.
Optimization techniques: whitespace reduction, deduplication, structure simplification.
Always create backup before applying optimization (backup: true).
Technical docs generation using Sphinx, MkDocs, TypeDoc, OpenAPI, and Nextra. Covers project configuration discovery, how to integrate with existing docs sites, and best practices for keeping framework documentation and technology guide content current.
| Generator | Use Case | Primary Format |
|---|---|---|
| Sphinx | Python projects, API docs | RST / Markdown |
| MkDocs | General projects | Markdown |
| TypeDoc | TypeScript libraries | TypeScript JSDoc |
| OpenAPI / Swagger | REST APIs | YAML / JSON |
| Nextra | Next.js docs sites | MDX |
.moai/config/sections/language.yaml/moai sync artifacts: README, CHANGELOG, API referenceJIT docs (just in time docs) — on-demand documentation discovery and loading based on user intent and conversation context.
, ): Official library docs.moai/, README, SPEC documents)Token budget: 5000 tokens per JIT load. Summarize if source exceeds budget.