Piensa 🧠

A simple CLI tool to interact easily with LLMs like OpenAI and Anthropic. Pipe in prompts, text, or files to quickly summarize, calculate, or get insights.

Installation

npm install -g piensa

Usage

Quick Examples

Ask a simple question:

piensa "What's the capital of France?"

Pipe a calculation:

echo "2 + 2" | piensa "calculate"

Summarize a file:

cat myfile.txt | piensa "summarize"

Stream responses as they're generated:

piensa "Tell me a story" --stream

Options

Use a specific provider or model:

piensa "Question?" --provider openai
piensa "Question?" --model gpt-4

Set your API key (stored for reuse):

piensa "Question?" --key your-api-key

Stream the response in real-time:

piensa "Question?" --stream
# or use the short form
piensa "Question?" -s

Config

Set up your preferences:

piensa --config

Check your current config:

piensa config

Set a default model:

piensa set-model openai gpt-4-turbo

Check your default model:

piensa get-model openai

Config Location

• macOS: ~/Library/Preferences/piensa-nodejs/config.json
• Linux: ~/.config/piensa-nodejs/config.json
• Windows: %APPDATA%\piensa-nodejs\Config\config.json

Security Note

⚠️ API Key Security: API keys are stored in plain text in the configuration file. While these files are protected by your operating system's user permissions, please be aware of the following risks:

• Do not share your configuration directory or backups containing these files
• Be cautious when using on shared computers
• Consider using environment variables for API keys in sensitive environments
• If your machine is compromised, an attacker could potentially access these keys

Supported Providers

• OpenAI (default)
• Anthropic

Development

Clone and run locally:

git clone https://github.com/cesarvarela/piensa.git
cd piensa
npm install
npm run build
npm start "Your prompt here"

License

ISC