Why does CodeRabbit return "Error during additional research" when fetching Jira tickets?
Last updated: July 10, 2026
Context
When CodeRabbit attempts to look up Jira issue details during a pull request review, you may encounter an Error during additional research message — even if your Jira integration appears connected and healthy in the CodeRabbit dashboard.
Answer
This error is caused by a connection-type mismatch. CodeRabbit fetches Jira ticket details during reviews via an MCP (Model Context Protocol) server connection. The standard native Jira integration (found under Integrations → Issue Tracking) supports issue/PR linking but does not support the research feature used during reviews. Reinstalling the native integration will not resolve this issue.
To fix this, you need to add Jira as an MCP server using Atlassian's Remote MCP endpoint, in addition to your existing native Jira integration. Follow the steps below:
Note: This setup must be performed by an Atlassian admin, who will also need to keep the OAuth connection active.
In CodeRabbit
Navigate to Integrations → MCP in the CodeRabbit dashboard.
Click New MCP server.
Fill in the following fields:
Server label:
Jira(or any name you prefer)Server URL:
https://mcp.atlassian.com/v1/sse(must be HTTPS — this is Atlassian's Remote MCP endpoint)Auth method: OAuth
Leave the access token and custom headers fields blank (OAuth handles authentication).
Click Save. CodeRabbit will open an OAuth popup to begin the authorization flow.
On the Atlassian/Jira side
Log in to Atlassian and select your site when prompted.
On the consent screen, click Accept/Authorize to grant CodeRabbit the requested scopes.
Ensure the Atlassian account being authorized has access to the relevant Jira projects and issues.
Once the MCP server connection is established, CodeRabbit will be able to fetch Jira ticket details during pull request reviews without errors.
For more information on setting up the native Jira integration, refer to the CodeRabbit Jira Integration documentation.