Quick Start

Get Zypher Agent from https://zypher.corespeed.io/, an open-source AI-agent framework/SDK within the CoreSpeed platform for building production-ready autonomous agents. This guide will help you get a Zypher Agent running in under 5 minutes.

Prerequisites

• Deno 2.0+ – install from https://docs.deno.com/runtime/getting_started/installation/
• API Keys
  • Gemini API Key – https://aistudio.google.com/
  • Firecrawl API Key – https://www.firecrawl.dev/
    • Gemini serves as an example LLM server.
    • Firecrawl serves as an example MCP server for web crawling.
  • Note: To retrieve the model name for Gemini, click on the Key listing and copy the cURL quickstart.

Step 1: Clone the Project

git clone <this-project-https-clone-url>
cd ZypherMindCLI

Step 2: Configure Environment Variables

Create a .env file in your project root:

GEMINI_API_KEY=your_gemini_api_key_here
FIRECRAWL_API_KEY=your_firecrawl_api_key_here

Step 3: Install VSCode Extension

Install the Deno VSCode extension from https://deno.land. This provides a language server client for Deno.

Step 4: Run Your Agent

Start your agent using:

deno run -A main.ts

• The -A flag grants all permissions (--allow-net --allow-read --allow-env, etc.) and automatically fetches and caches dependencies.

Once running, the terminal interface allows you to:

• Chat with your agent interactively
• Receive real-time streamed responses
• Ask follow-up questions within the same session

Try Different Tasks

After your agent is running, you can experiment with:

• Write a short motivational message for a startup team using bullet points.
• Rewrite the motivational message in a humorous tone and limit it to 3 bullets.
• Use Firecrawl to crawl https://openai.com/research and extract titles of the most recent research updates.
• etc.

About Feature Coverage

ZypherMindCLI focuses on the core capabilities best supported by the free-tier Gemini API: text generation and reasoning. These features match the project's goal of providing a lightweight, CLI-driven AI agent experience.

For a complete list of Gemini API features and their potential usage, see the official documentation: Gemini API What's Next.

Gemini API Feature Coverage in ZypherMindCLI

Already Supported:

• Text Generation – Full support using generateContent + streaming.
• Thinking / Reasoning – Automatically leveraged by Gemini models.

Available in Gemini, Not Implemented in This Project:

• Image Generation – Not added because CLI cannot display images.
• Image Understanding (Vision) – Not added because CLI does not handle image inputs.
• Function Calling – Optional; can be added later for complex agent tool orchestration.
• Long Context – Free-tier context window is limited; not needed for CLI.
• Embeddings – Not added; used mainly for RAG or search systems, outside this project's scope.

These advanced features can be implemented if you:

• Upgrade your Gemini API plan, or
• Want the agent to support richer capabilities beyond the CLI environment.

For most CLI-based use cases, the current implementation is optimal—simple, fast, and aligned with the strengths of the free-tier Gemini models.