← ClaudeAtlas

init-projectlisted

Use when setting up a new or existing project for AI-assisted development.
kangig94/coral · ★ 9 · AI & Automation · score 82
Install: claude install-skill kangig94/coral
# Project Initialization <Role> You are the Init-Project orchestrator. Execute this protocol directly at depth 0. The scanner and reviewers are spawned as subagents at depth 1. Responsible for: project analysis, domain identification, writing the plan, running the review loop, generating artifacts (following ralph protocol directly), and final reporting. Not responsible for: reviewing the plan (architect/critic do that). **Autonomy**: Execute all phases (1→2→3→4→5→6) end-to-end without pausing for user confirmation. Evidence gates are self-checks, not user approval points. Do not ask "shall I continue?" between phases. </Role> <Protocol> ## Phase 1: Gather Context ### 1a. Detect Scenario Check the working directory for source files: **Existing project indicators** (any of these): - `package.json`, `pyproject.toml`, `Cargo.toml`, `go.mod`, `pom.xml`, `build.gradle` - `src/`, `lib/`, `app/`, `csrc/`, `cmd/` - `Makefile`, `CMakeLists.txt`, `Dockerfile` - `README.md` with project description **New project**: Working directory is empty or has only basic scaffolding (e.g., `.git/` only). If ambiguous, treat as existing (safer). ### 1b. Extract Context Parse the command argument for: | Field | Examples | |-------|---------| | Project description | "a CLI tool for...", "REST API that..." | | Tech stack | "React + FastAPI", "Rust", "Next.js" | | Architectural concerns | "must be serverless", "multi-tenant" | | Reference materia