Skip to main content

​
Agentica SDK

The Agentica SDK is a type-safe agent framework that lets LLM agents integrate with your code — functions, classes, live objects, even entire SDKs. Instead of building MCP wrappers or brittle schemas, you pass references directly; the framework enforces your types at runtime, constrains return types, and manages agent lifecycle. Compose single agentic functions into multi-agent systems that feel like ordinary code and ship anywhere your stack runs.

​
Why use the Agentica SDK?

Direct SDK Integration

Pass real SDK clients or extract just the methods you need — no MCP servers, no schema work. Inject live objects into an agent’s scope with bindings and runtime state intact.

Type-Safe Agents

Function signatures become runtime contracts. Python annotations and TypeScript generics drive structured validation, reducing hallucinations and turning agentic workloads into predictable, testable operations.

Multi-Agent Orchestration

Spin up agents on demand, keep state across calls, stream output, and compose leader/worker patterns like regular functions. Standard patterns like initialization, closures, and automatic cleanup make complex systems simple.

Multi-Language Support

Models interact with your codebase in a consistent, unified way — regardless of language or runtime. Python and TypeScript SDKs available now, with more languages coming soon.

​
Interact with the Docs

If you have uv installed, you can interact with the Agentica SDK docs by chatting with an agent built with the Agentica SDK! Run uvx symbolica-agentica to start chatting with the agent, and ask it questions about the Agentica SDK! (requires an API key from here) 
╭─────────────────────────────────────────────────╮
│ ▣  Welcome to the Interactive Agentica SDK Docs │
╰─────────────────────────────────────────────────╯

* Using model: anthropic/claude-opus-4.6

───────────────────────────────────────────────

> Introduce yourself

Hey! I'm an agent created with the Agentica framework.

Would you like to see me demonstrate any of the following?

A: Tool use without MCP
B: Multi agent orchestration
C: How I have been defined and created with the Agentica SDK?

or I can answer any questions you may have about the
Agentica framework.

> Type your message...

​
Quick Example

Here’s how simple it is to create an agent-backed function: 
from agentica import agentic
from typing import Literal

@agentic()
async def analyze(text: str) -> Literal["positive", "neutral", "negative"]:
    """Analyze sentiment"""
    ...

result = await analyze("The Agentica SDK is an awesome framework!")

​
Getting Started

Ready to build your first Agentica SDK agent? Follow our quickstart guide to get up and running in minutes.

Quickstart Guide

Install the Agentica SDK and build a complete note-taking assistant with agentic functions and agents. Supports Python and TypeScript.
TypeScript Framework Support
  • Plain TS/JS (Node, Bun, Browser) via ts-patch
  • Next.JS
  • React
  • Vite
  • Svelte
  • Webpack
  • Rollup
  • ESBuild
After installing the Agentica SDK, create an .env file with your API key:
echo 'VITE_AGENTICA_API_KEY=<YOUR_KEY>' >> .env
Client-side usageVite requires the VITE_ prefix to expose variables to client code. Note that this exposes the key to the browser - only do this if your API key is meant for client-side use.

​
Explore the Documentation

Core Concepts

Understand the fundamentals of how the Agentica SDK works, scope, and when to use agentic functions vs agents.

Guides

Task-oriented tutorials for multi-agent systems, error handling, and best practices.

Examples

Complete working applications including Slack bots, data analysis, and research systems.

API Reference

Complete API documentation for Python and TypeScript with all parameters and types.

Copy the following markdown as context for Cursor, Claude Code, or Copilot:
See our integration guides for more details.

Need help? Contact us via the button below!