Forra Developer Guide - Getting Started

Forra is a comprehensive AI platform that provides developers with powerful tools and SDKs to build, integrate, and extend AI-powered applications. Whether you're creating custom assistants, building chat experiences, or developing specialized workflows, Forra has the tools you need.

What Forra Offers Developers

Forra provides a complete ecosystem for AI development:

🤖 Custom AI Assistants: Create specialized assistants with Skills and workflows
💬 Chat Integration: Embed Forra's chat interface into your applications
🔧 Skills: Extend assistants with Python-based functions and custom logic
🎨 Applications: Build custom frontend interfaces for specialized user experiences
📡 REST API: Direct API access for maximum flexibility
🛠️ Development Tools: CLI tools for project initialization, packaging, and deployment

Development Paths

1. Using Forra as an LLM

Drop-in replacement for OpenAI or Anthropic in your existing tools and workflows

Use Forra with any OpenAI- or Anthropic-compatible library or framework - no code changes needed. Simply point your existing SDK calls to Forra's endpoints.

OpenAI Compatibility Guide - Use Forra with existing OpenAI tools and frameworks
Anthropic Compatibility Guide - Use Forra with Claude Code and Anthropic-compatible tools

2. Integrating Forra

Build applications that leverage Forra's capabilities

Integrate Forra into your applications using REST API or language-specific SDKs. Create conversations, call assistants, manage users, and embed chat interfaces.

REST API Documentation - Complete HTTP API reference
Python SDK Getting Started - Setup and basic integration with Python
JavaScript API SDK - Direct API client for TypeScript/JavaScript
React Chat SDK - Pre-built React chat components

3. Extending Forra

Add custom capabilities to your assistants

Extend Forra assistants with custom functionality using Skills (Python functions, webhooks) and Applications (custom frontend interfaces).

Skills Overview - Understanding Skills, scopes, and types
Functions Getting Started - Create your first functions
Applications Getting Started - Build custom UI experiences

Core Concepts

Assistants

The foundation of Forra - AI agents that can be customized with: - Skills (reusable functionality that can be private or shared) - Knowledge bases (document uploads) - Applications (custom UIs) - Configuration variables and secrets

Conversations

Sessions between users and assistants that maintain context and allow for: - Multi-turn dialogues - State persistence through user_data - File uploads and downloads - Custom workflow tracking

Skills

Reusable logic that extends assistant capabilities. Skills can be: - Global Skills: Shared across multiple assistants, managed by admins - Assistant Skills: Specific to a single assistant

Functions are the primary implementation type for Skills:

from forrasdk import forra
from pydantic import Field

@forra.function(description="Calculate compound interest")
def calculate_interest(principal: float = Field(description="Amount to be invested"), rate: float = Field(description="Interest rate"), time: int = Field(description="Time invested")):
    return principal * (1 + rate) ** time

Learn more about Skills →

Applications

Custom frontend interfaces that provide specialized user experiences: - Built with any web framework - Access to Forra's API through provided tokens - Can manage conversations and execute functions - Deployed as zipped web applications

Ready to build with Forra? Pick your development path above and dive into the detailed documentation!