Does Agentica support Model Context Protocol (MCP)?

Yes, Agentica has native MCP integration, allowing you to connect agents to MCP servers via configuration files for expanded tool capabilities.

Can I track token usage with this skill?

Yes, the SDK provides 'last_usage()' and 'total_usage()' methods to monitor input, output, and cumulative token processing for cost management.

How do I handle persistent memory in Agentica?

For @agentic functions, you can set 'persist=True' to remember conversation history. For agents created via spawn(), state is automatically maintained across calls to the same instance.

What is Agentica SDK?

Agentica is a Python framework designed for building AI agents that can maintain state, implement functions, use external tools, and coordinate with other agents.

Which LLM models can I use with Agentica?

Agentica supports major models from Anthropic (Claude), OpenAI (GPT-4/5), and any provider available through OpenRouter slugs like Google Gemini.

Agentica SDK

Name: Agentica SDK
Author: scooter-lacroix

byscooter-lacroix

0•

Data Science & ML

Builds and orchestrates stateful Python-based AI agents with MCP integration and multi-agent patterns.

The Agentica SDK skill provides a comprehensive framework for developing sophisticated AI agents in Python. It enables developers to transform standard functions into AI-powered tools using the @agentic decorator, spawn autonomous agents with specific premises, and manage complex multi-agent orchestrations. With built-in support for the Model Context Protocol (MCP), token usage tracking, conversation persistence, and OpenTelemetry tracing, this skill is essential for building production-grade AI applications that require structured outputs, tool-use, and robust state management.

Key Features

01Seamless Model Context Protocol (MCP) integration for connecting external tools

02Asynchronous agent spawning with custom system prompts and model selection

03Support for multi-agent coordination and persistent state across conversation history

04Advanced token management including usage tracking and fine-grained limits

05@agentic decorator for creating AI-powered Python functions with minimal boilerplate

060 GitHub stars

Use Cases

01Developing complex multi-agent workflows for automated research or data analysis

02Integrating external APIs and legacy tools into LLM-driven applications via MCP

03Building stateful chatbots and assistants that maintain context across multiple interactions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add scooter-lacroix/maestro agentica-sdk

For use in Claude.ai and ChatGPT

Download Skill