I am creating an AI agent that interacts with Jira via MCP (Model Context Protocol) using https://mcp.atlassian.com/v1/mcp
I followed these steps:
Set up Azure OpenAI LLM - Configure an Azure ChatOpenAI model using credentials from your .env file (endpoint, API key, deployment name)
Connect to Atlassian's Rovo MCP server — Authenticated with mcp rovo server using Basic Authentication with email and an API token (this is an scoped API token for Rovo MCP), then fetch all available MCP tools (this is succeeding - I am getting list of tools)
Creating a ReAct agent — LangChain's createReactAgent combined with the Azure LLM with the Jira tools so the AI can reason about which tools to call
Executed this query — You invoke the agent with the message: "Give me jira issue jira-1. And my current account Cloud ID is https://company.atlassian.net"
Results - ToolException: MCP tool 'getJiraIssue' on server 'atlassian' returned an error: {"error":true,"message":"You
don't have permission to connect via API token. Please ask your organization admin for access."}
The error indicates you lack API token access, likely due to restrictions set by your Org Admins in the authentication policy. You will have to contact your Org Admin or Jira Admin team for assistance.
Hi @Varkha Sharma - welcome to the Community,
are you the Organisation Admin? The error message points to API token settings, but I don't think it has anything to do with Authentication policies (that would prohibit you from creating a token) but rather the MCP settings.
You (or the Org Admin) need go to admin.atlassian.com > Rovo > Rovo MCP server > Authentication > Allow API token authentication > Enable
Could you confirm if that setting helps?
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.