minicode-sdk Usage Guide

Installation

pip install minicode-sdk

Quick Start

Basic Agent

import asyncio
from minicode import Agent
from minicode.llm import OpenAILLM

async def main():
    agent = Agent(
        name="assistant",
        llm=OpenAILLM(api_key="your-api-key"),
    )

    response = await agent.generate("Hello, how are you?")
    print(response)

asyncio.run(main())

Streaming Response

async def main():
    agent = Agent(name="assistant", llm=OpenAILLM(api_key="your-key"))

    async for chunk in agent.stream("Tell me a story"):
        if chunk["type"] == "content":
            print(chunk["content"], end="", flush=True)

Adding Tools

Built-in Tools

from minicode.tools.builtin import ReadTool, WriteTool, BashTool

agent = Agent(
    name="assistant",
    llm=my_llm,
    tools=[ReadTool(), WriteTool(), BashTool()],
)

Custom Tools

from minicode.tools.base import BaseTool

class MyTool(BaseTool):
    @property
    def name(self) -> str:
        return "my_tool"

    @property
    def description(self) -> str:
        return "Description of what this tool does"

    @property
    def parameters(self) -> dict:
        return {
            "type": "object",
            "properties": {
                "param1": {"type": "string", "description": "First parameter"},
            },
            "required": ["param1"],
        }

    async def execute(self, params: dict, context) -> dict:
        # Implement tool logic
        return {"success": True, "result": "..."}

MCP Integration

Method 1: Direct Configuration

async with Agent(
    name="assistant",
    llm=my_llm,
    mcp_servers=[
        {
            "name": "memory",
            "command": ["npx", "-y", "@modelcontextprotocol/server-memory"],
        }
    ],
) as agent:
    response = await agent.generate("Remember that my name is Alice")

Method 2: Configuration File

Create .minicode/mcp.json:

{
  "mcpServers": {
    "memory": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-memory"]
    }
  }
}

Then use the agent (MCP servers are loaded automatically):

async with Agent(name="assistant", llm=my_llm) as agent:
    # MCP tools are automatically available
    pass

Agent Instructions

Create .minicode/AGENT.md to define custom instructions:

# Project Guidelines

- Use Google-style docstrings
- All code must be production-ready
- Ask for clarification if requirements are unclear

Instructions are automatically injected into user messages.

Disable with:

agent = Agent(
    name="assistant",
    llm=my_llm,
    use_agent_instructions=False,
)

Skills

Using SkillTool

from minicode.tools.builtin import SkillTool

agent = Agent(
    name="assistant",
    llm=my_llm,
    tools=[SkillTool()],
)

Creating Skills

Create .minicode/skills/my-skill/SKILL.md:

---
name: my_skill
description: What this skill does and when to use it.
---

# Skill Content

Instructions for the agent when this skill is invoked.

Configuration Locations

Environment Variables