Using Composio With IBM

Integrate Composio with IBM’s Granite Models to let them seamlessly interact with external apps

Composio enables IBM’s Granite Models to connect with many tools!

Goal: Star a repository on GitHub with natural language

Video Guide

Install Packages & Connect a Tool

Ensure you have the necessary packages installed and connect your GitHub account to allow your IBM agents to utilize GitHub functionalities.

Run command

pip install composio-langchain
pip install langchain-ibm

#  Connect your GitHub so agents can use it. 
composio add github
#  Check all different apps which you can connect with
composio apps

Goal: Prepare your environment by initializing necessary imports from Composio & IBM.

1. Import Base Packages

   Prepare your environment by initializing necessary imports from Composio & IBM.

   Default Imports

   from composio_langchain import ComposioToolSet, Action
   from langchain_ibm import ChatWatsonx
   import os
   
   os.environ['WATSONX_API_KEY'] = '<ibm_api_key>' #add your ibm api key here
   if not os.environ.get('WATSONX_API_KEY'):
       raise ValueError("WATSONX_API_KEY environment variable is not set")
   
2. Integrating GitHub Tools with Composio

   This step involves fetching and integrating GitHub tools provided by Composio, enabling enhanced functionality for Agent operations.

   IBM Supported tools from Composio

   composio_toolset = ComposioToolSet()
   tools = composio_toolset.get_tools(
       actions=[Action.GITHUB_ACTIVITY_STAR_REPO_FOR_AUTHENTICATED_USER]
   )
   
3. IBM Agent Setup

   This step involves configuring and executing the agent to carry out actions, such as starring a GitHub repository.

   IBM Agent Setup

   parameters = {
   "decoding_method": "sample",
   "max_new_tokens": 100,
   "min_new_tokens": 1,
   "temperature": 0.5,
   "top_k": 50,
   "top_p": 1,
   }
   url = input('Add your IBM Cloud URL here: ')
   project_id = input('Add your IBM Project ID here: ')
   watsonx_llm = ChatWatsonx(
   model_id = 'ibm/granite-3-8b-instruct',
   url = url, 
   project_id = project_id,
   )   
   
   if not url or not project_id:
       raise ValueError("IBM Cloud URL and Project ID must be provided")
   
   llm_with_tools = watsonx_llm.bind_tools(tools)
   
   
   
4. Execute with your prompt/task.

   Execute the following code to execute the agent, ensuring that the intended task has been successfully completed.

   Agent Execution

   response = llm_with_tools.invoke("Star the composiohq/composio repository")
   print(response)