This MCP server combines two toolkit integrations to extend the IBM WatsonX Orchestrate ADK Salesforce agent: Tavily, an AI-powered web search toolkit, and Atlassian Jira tooling for issue tracking and project management. After importing the Tavily and Atlassian MCP toolkits, the agent gains access to a suite of tools for researching external information, extracting content from URLs, and performing Jira operations such as querying issues, creating issues, and updating or commenting on issues. You can invoke these tools from within Salesforce leads, contacts, or custom workflows to enrich data with external knowledge, verify information, or automate project-related tasks. The Tavily tools enable online research and content extraction, while the Jira tools cover issues, projects, sprints, and workflows, enabling a more connected sales and support workflow with up-to-date context from external sources.

Prerequisites:

• Python 3.8+
• IBM watsonx Orchestrate ADK installed
• Salesforce account with API access

Installation steps:

1. Clone the repository and install dependencies: git clone https://github.com/IBM/orchestrate-adk-agent.git cd orchestrate-adk-agent pip install -r requirements.txt

2. Create and configure environment variables from the template: cp .env.template .env

   Edit .env and populate Salesforce, Tavily, and Jira credentials as needed

3. Ensure shell scripts have execute permissions (if you plan to run quick setup): chmod +x import_agent.sh setup_connection.sh check_connection.sh test_agent.sh

4. Import MCP toolkits (as described in the MCP sections of the README):

   Tavily MCP toolkit import:

   ./import_mcp_toolkit.sh

   Atlassian MCP toolkit import:

   ./import_atlassian_mcp.sh

5. Verify toolkits are listed: orchestrate toolkits list

6. Run tests or proceed to build and run the agent within your environment as described in the repository.

Notes and tips:

• The Tavily MCP toolkit relies on a Tavily API key stored in a dedicated connection (tavily_creds). Configure the connection and environment as directed by the setup scripts.
• The Atlassian Jira toolkit requires a Jira Personal Access Token and a Jira URL. Use the provided quick setup scripts or the manual import flow to configure the connection.
• Ensure you have Node.js installed if you plan to use the Tavily npx-based toolkit import, and ensure uvx is available for the Atlassian toolkit import.
• When using env vars, store sensitive credentials in the orchestrate connections/credentials management workflow rather than hard-coding in scripts.
• If you encounter import or authentication errors, re-run the respective import scripts and verify that the token scopes align with the required API access.