AgentMind Plugin Marketplace

Welcome to the AgentMind Plugin Marketplace! This directory contains information about available plugins and how to create your own.

Available Plugin Types

AgentMind supports the following plugin types:

LLM Providers - Custom language model integrations
Memory Backends - Storage backends for agent memory
Tools - External tools and APIs agents can use
Orchestrators - Custom orchestration strategies
Observers - Monitoring and middleware plugins

Official Plugins

LLM Providers

agentmind-plugin-openai - OpenAI GPT models integration
agentmind-plugin-anthropic - Anthropic Claude integration
agentmind-plugin-ollama - Local Ollama models (built-in)
agentmind-plugin-litellm - LiteLLM unified interface (built-in)

Memory Backends

agentmind-plugin-redis - Redis-based memory backend
agentmind-plugin-pinecone - Pinecone vector database
agentmind-plugin-weaviate - Weaviate vector database
agentmind-plugin-chroma - ChromaDB integration (built-in)

Tools

agentmind-plugin-web-search - Web search capabilities
agentmind-plugin-code-executor - Safe code execution
agentmind-plugin-file-ops - File system operations
agentmind-plugin-api-client - REST API client

Orchestrators

agentmind-plugin-langgraph - LangGraph compatibility
agentmind-plugin-autogen - AutoGen-style orchestration
agentmind-plugin-crewai - CrewAI-style workflows

Observers

agentmind-plugin-langsmith - LangSmith tracing
agentmind-plugin-wandb - Weights & Biases logging
agentmind-plugin-prometheus - Prometheus metrics

Installing Plugins

Install plugins using pip:

pip install agentmind-plugin-openai
pip install agentmind-plugin-redis

Plugins are automatically discovered via entry points. No additional configuration needed!

Using Plugins

from agentmind import discover_plugins, load_plugin

# Discover all available plugins
plugins = discover_plugins()
print(plugins)
# {'llm': ['openai', 'anthropic'], 'memory': ['redis', 'pinecone'], ...}

# Load a specific plugin
OpenAIProvider = load_plugin('llm', 'openai')
provider = OpenAIProvider(api_key='your-key')

# Use with AgentMind
from agentmind import Agent, AgentMind

agent = Agent(name="assistant", llm_provider=provider)
mind = AgentMind(llm_provider=provider)

Creating Your Own Plugin

1. Create Plugin Package Structure

agentmind-plugin-myplugin/
├── setup.py
├── README.md
├── LICENSE
├── agentmind_plugin_myplugin/
│   ├── __init__.py
│   └── plugin.py
└── tests/
    └── test_plugin.py

2. Implement Plugin Interface

# agentmind_plugin_myplugin/plugin.py
from agentmind.plugins.interfaces import LLMProvider, PluginMetadata

class MyLLMProvider(LLMProvider):
    """My custom LLM provider."""
    
    def get_metadata(self) -> PluginMetadata:
        return PluginMetadata(
            name="myplugin",
            version="0.1.0",
            description="My custom LLM provider",
            author="Your Name",
            plugin_type="llm_provider",
        )
    
    async def initialize(self, config=None):
        # Initialize your provider
        pass
    
    async def shutdown(self):
        # Cleanup
        pass
    
    def health_check(self):
        return True
    
    async def generate(self, messages, **kwargs):
        # Implement generation logic
        return {
            "content": "Generated response",
            "model": "my-model",
            "usage": {},
            "metadata": {},
        }
    
    async def generate_stream(self, messages, **kwargs):
        # Implement streaming
        yield "chunk1"
        yield "chunk2"
    
    def get_model_info(self):
        return {"name": "my-model", "max_tokens": 4096}

3. Configure Entry Points

# setup.py
from setuptools import setup, find_packages

setup(
    name="agentmind-plugin-myplugin",
    version="0.1.0",
    description="My AgentMind plugin",
    author="Your Name",
    author_email="your.email@example.com",
    packages=find_packages(),
    install_requires=[
        "agentmind>=0.2.0",
        # Your dependencies
    ],
    entry_points={
        "agentmind.plugins.llm": [
            "myplugin = agentmind_plugin_myplugin.plugin:MyLLMProvider",
        ],
    },
    classifiers=[
        "Development Status :: 3 - Alpha",
        "Intended Audience :: Developers",
        "License :: OSI Approved :: MIT License",
        "Programming Language :: Python :: 3.8",
        "Programming Language :: Python :: 3.9",
        "Programming Language :: Python :: 3.10",
        "Programming Language :: Python :: 3.11",
    ],
)

4. Test Your Plugin

# tests/test_plugin.py
import pytest
from agentmind_plugin_myplugin.plugin import MyLLMProvider

@pytest.mark.asyncio
async def test_plugin():
    provider = MyLLMProvider()
    await provider.initialize()
    
    result = await provider.generate([
        {"role": "user", "content": "Hello"}
    ])
    
    assert result["content"]
    assert provider.health_check()
    
    await provider.shutdown()

5. Publish Your Plugin

# Build distribution
python setup.py sdist bdist_wheel

# Upload to PyPI
pip install twine
twine upload dist/*

Plugin Templates

AgentMind provides templates for all plugin types:

from agentmind.plugins.templates import (
    ExampleLLMProvider,
    ExampleMemoryBackend,
    ExampleOrchestrator,
    ExampleObserver,
    print_plugin_template,
)

# Print full template
print_plugin_template()

Plugin Guidelines

Best Practices

Follow the interface - Implement all required methods
Handle errors gracefully - Don't crash the main application
Document thoroughly - Provide clear documentation and examples
Test extensively - Include comprehensive tests
Version properly - Use semantic versioning
Declare dependencies - List all requirements in setup.py

Security

Validate inputs - Never trust user input
Use sandboxing - For code execution plugins
Respect permissions - Declare required permissions
Secure credentials - Never hardcode API keys
Audit dependencies - Keep dependencies up to date

Performance

Async by default - Use async/await for I/O operations
Cache when possible - Reduce redundant operations
Batch operations - Group similar operations
Resource cleanup - Properly close connections
Monitor usage - Track resource consumption

Community Plugins

Submit your plugin to be listed here! Create a PR with:

Plugin name and description
Installation instructions
Link to repository
License information

Support

Documentation: https://agentmind.readthedocs.io
GitHub Issues: https://github.com/cym3118288-afk/AgentMind-Framework/issues
Discord: [Join our community]
Email: cym3118288@gmail.com

License

All official plugins are MIT licensed. Community plugins may have different licenses.

FilesExpand file tree

plugin_marketplace.md

Latest commit

History