Agent Skills › evolution-foundation/evo-nexus › dev-deepinit

dev-deepinit

GitHub

为新项目或重构后项目生成分层上下文文档（AGENTS.md/CLAUDE.md），帮助工程智能体快速掌握代码库结构、技术栈及规范，避免重复探索。

.claude/skills/dev-deepinit/SKILL.md evolution-foundation/evo-nexus

Trigger Scenarios

新项目初始化缺乏上下文文档的项目重大重构后

Install

npx skills add evolution-foundation/evo-nexus --skill dev-deepinit -g -y

More Options

Non-standard path

npx skills add https://github.com/evolution-foundation/evo-nexus/tree/main/.claude/skills/dev-deepinit -g -y

Use without installing

npx skills use evolution-foundation/evo-nexus@dev-deepinit

指定 Agent (Claude Code)

npx skills add evolution-foundation/evo-nexus --skill dev-deepinit -a claude-code -g -y

安装 repo 全部 skill

npx skills add evolution-foundation/evo-nexus --all -g -y

预览 repo 内 skill

npx skills add evolution-foundation/evo-nexus --list

SKILL.md

Frontmatter

{
    "name": "dev-deepinit",
    "description": "Deep codebase initialization — generate hierarchical AGENTS.md \/ CLAUDE.md context files for a new project so engineering agents have full context from session start."
}

Dev Deepinit

Derived from oh-my-claudecode (MIT, Yeachan Heo). Adapted for the EvoNexus Engineering Layer.

Deep codebase initialization. For a new project, generate hierarchical AGENTS.md / CLAUDE.md context files that give engineering agents full project context from session start — instead of forcing them to re-discover the codebase every session.

Use When

Onboarding a new project to EvoNexus
Existing project lacks AGENTS.md / CLAUDE.md / project context docs
After a major refactor that invalidates existing context docs

Do Not Use When

Project already has comprehensive context docs
Trivial / experimental codebases (overhead not worth it)

Workflow

Phase 1 — Discovery

Delegate to @scout-explorer:

Map the directory structure
Identify the language(s), framework(s), build tools
Find entry points (main files, package scripts, CI configs)
Find test setup
Find existing docs (README, ARCHITECTURE.md, etc.)

Phase 2 — Synthesis

Project purpose (1-2 sentences)
Tech stack
Architecture overview
Key directories with their purpose
Build / test / run commands
Conventions detected (naming, error handling, imports)
Known gotchas (from comments, TODOs, recent commits)

Phase 3 — Generation

Create the hierarchical context:

{project-root}/
├── CLAUDE.md          # Top-level project context (this is what Claude reads first)
├── AGENTS.md          # Detailed context for engineering agents
└── docs/
    └── architecture/
        └── [C]project-overview-{date}.md

Phase 4 — Verification

@oath-verifier reads the new context docs and confirms they're accurate

Output Files

CLAUDE.md (top of project)

# {Project Name}

[1-2 sentence purpose]

## Tech Stack
- Language: ...
- Framework: ...
- Build: ...

## Quick Commands
- Build: `...`
- Test: `...`
- Run: `...`

## Architecture
[high-level overview]

## Conventions
[detected patterns]

AGENTS.md (detailed context)

# Engineering Context

## Module Map
- `src/auth/` — authentication
- `src/db/` — database layer
- ...

## Key Files
- `src/main.ts:42` — entry point
- ...

## Patterns
[deep convention notes]

## Gotchas
[known pitfalls]