PE2-CLI

Command-line tool that takes a rough prompt (text or file), calls a configured LLM, and returns a structured PE2-style prompt. Release line is 4.0.0 (npm semver); the CLI labels the generation as Code V4 in the banner.

Requirements

• Node.js 18.17 or newer

You need an API key (or base URL for Ollama) for whichever provider you pick. The tool does not ship bundled models.

Install

npm install -g @kleosr/pe2-cli

Or run without a global install:

npx @kleosr/pe2-cli --help

Usage

No arguments starts interactive mode (and will ask for config on first run if nothing is saved):

npx @kleosr/pe2-cli

Other entry points:

npx @kleosr/pe2-cli --config
npx @kleosr/pe2-cli "Your prompt as plain text"
npx @kleosr/pe2-cli path/to/prompt.txt
npx @kleosr/pe2-cli "Some text" --iterations 3
npx @kleosr/pe2-cli "Some text" --provider ollama --model llama3
npx @kleosr/pe2-cli "Some text" --auto-difficulty

For the full flag list, use:

npx @kleosr/pe2-cli --help

Configuration

Settings are stored in ~/.kleosr-pe2/config.json (Unix) or the equivalent under your user profile on Windows. Run --config to change provider, model, and API key.

Supported providers in code: openai, anthropic, google, openrouter, ollama. Default in the project constants targets OpenRouter with a small model id; you can override per run with --provider and --model.

Output

Unless you pass --output-file, session output can be written under pe2-prompts/ in the current working directory (see src/paths.js and src/engine.js).

Development

git clone https://github.com/kleosr/PE2-CLI.git
cd PE2-CLI
npm ci
npm test
npm start

Tests use Node’s built-in runner: node --test ./tests/*.test.js.

Package

• npm: @kleosr/pe2-cli
• Source: github.com/kleosr/PE2-CLI

License

ISC (see package.json).