# Init Project

> Full project initialization workflow from API key setup to GitHub repo creation and issue creation Use when this capability is needed.

- Skill: `tomevault-io/init-project-3` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add tomevault-io/init-project-3`
- Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/init-project-3/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Integrations & APIs
- Author: tomevault-io (https://skillmd.com/u/tomevault-io)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/tomevault-io/init-project-3

---


# init-project Skill

## Purpose
Initialize a new project with complete planning and setup workflow.

## When to Use
- Starting a new project from scratch
- Setting up a fork of the power_template for a new codebase
- When you want a complete, AI-assisted project initialization

## Workflow

### Phase 1: Setup and Verification

#### 1.1 Check Environment Variables
Verify required environment variables are set:

```bash
echo $OPENROUTER_API_KEY
echo $GEMINI_API_KEY
echo $GITHUB_TOKEN
```

If any are missing, guide the user to set them:

```bash
# Set in ~/.bashrc or ~/.zshrc for persistence
export OPENROUTER_API_KEY="your-key-here"
export GEMINI_API_KEY="your-key-here"
export GITHUB_TOKEN="your-token-here"
```

#### 1.2 Verify GitHub Authentication
```bash
gh auth status
```

If not authenticated:
```bash
gh auth login
```

### Phase 2: Project Planning

#### 2.1 Gather Requirements

**FIRST INTERACTION - FREE TEXT ONLY**
Ask the user: "What project do you have in mind?"
- DO NOT provide any pre-defined options or multiple-choice answers
- Use free text input only - let the user describe their project in their own words
- Wait for the user's response before proceeding

**AFTER USER RESPONSE**
Then ask clarifying questions to understand:
- What problem are you solving?
- Who are the users?
- What are the main features?
- Any functional requirements?
- Any non-functional requirements (latency, scale, cost, security)?
- Any constraints or limitations?
- Any preferences for languages, frameworks, databases, deployment?

#### 2.2 Create Technical Design
Use Planner Agent to create `docs/technical-design.md` with:
- Project overview
- Architecture
- Data model
- API design
- Security considerations
- Deployment strategy
- Define subsystem boundaries (e.g., frontend, backend, api, database, infrastructure)

#### 2.3 Initialize Version Tracking
Use Planner Agent to initialize subsystem version tracking:
- Create `docs/versions.md` with all subsystems tracked
- Determine appropriate version file location for each subsystem:
  - JavaScript/TypeScript: Use `package.json` version field
  - Python: Use `pyproject.toml` or `__version__.py`
  - Java: Use `pom.xml` or `build.gradle`
  - Go: Use module version or create `version.go`
  - API: Use `openapi.yaml` info.version field
  - Database: Create `migrations/version.json`
  - Infrastructure: Create `terraform/version.json` or similar
- Initialize all subsystems to version 1.0.0 in their respective locations
- Create version files where needed (e.g., `api/version.json`, `migrations/version.json`)

#### 2.4 Create Execution Plan
Use Planner Agent to create `docs/execution-plan.md` with:
- Task list with priorities
- Dependencies
- Milestones
- Acceptance criteria

### Phase 3: GitHub Setup (Semi-Automated)

#### 3.1 Propose Repository Configuration
Generate proposed repository configuration:
- Repository name (based on project)
- Description
- Visibility (private by default)
- Topics

#### 3.2 Get User Confirmation
Ask user to:
- Confirm proposed configuration
- Or modify as needed

#### 3.3 Create Repository
```bash
gh repo create $REPO_NAME \
  --description "$DESCRIPTION" \
  --$VISIBILITY \
  --topics "$TOPICS"
```

#### 3.4 Set Up Gemini Code Assist
Before proceeding with development:
1. Install Gemini Code Assist from https://github.com/apps/gemini-code-assist
2. Select the newly created repository
3. Grant necessary permissions for code reviews

#### 3.4 Create GitHub Issues
Parse execution plan and create issues:
- One issue per task
- Include detailed instructions
- Add dependencies between issues

### Phase 4: Initialize Project Documentation

#### 4.1 Create Project-Specific README
Use Builder Agent to create a new `README.md` for the project:
- Project name and description
- Quick start instructions
- Features overview
- Installation guide
- Usage examples
- Development instructions
- Contributing guidelines
- License information

Replace the template README with project-specific content.

#### 4.2 Commit Initial Setup
Commit the initial setup:
```bash
git add -A
git commit -m "chore: initialize project with [project-name]

- Set up operational state tracking
- Create technical design and execution plan
- Initialize subsystem version tracking
```

#### 4.3 Push to Remote
```bash
git push -u origin main
```

## Usage

```bash
/skill init-project
```

Or with shorter command:
```bash
/init
```

## Required Environment Variables
- `OPENROUTER_API_KEY` - For GLM4.5 Air and Gemini2.0 models
- `GEMINI_API_KEY` - For web searching and documentation
- `GITHUB_TOKEN` - For repository creation and issue management

## Required Agents
- `planner` - Creates technical design and execution plan
- `web_searcher` - Researches best practices and options

## Configuration
This skill uses the configuration from `.git/opencode`:
- Model providers (OpenRouter, Gemini)
- Agent configurations
- Skill-specific settings

---
> Converted and distributed by [TomeVault](https://tomevault.io/claim/thiagorcdl) — claim your Tome and manage your conversions.
<!-- tomevault:4.0:skill_md:2026-04-15 -->