Skip to main content
This feature is currently in open beta. We are actively improving it based on your feedback. If you encounter any issues or have suggestions, please share them on our Discord community or reach out to our support team.

ā€‹
Prerequisites

Enable the Jira integration as explained in the integrations documentation
Existing Jira users: If you enabled Jira before this feature was released, youā€™ll need to re-authenticate to grant the additional permissions required for planning.

ā€‹
Initiating Planning

ā€‹
Manual Planning

CodeRabbit Command

Comment @coderabbitai plan on any issue to generate a plan.
Navigate to the Planning tab in the CodeRabbit web app to configure automatic planning rulesets.
Configuration via Web UI Only: Auto-planning rulesets must be configured through the CodeRabbit web interface. Configuration via .coderabbit.yaml is not supported for this feature.
A ruleset consists of conditions that, when met, automatically trigger plan generation. All conditions are optionalā€”you can use any combination that fits your workflow:
ConditionDescription
Issue TypeMatch specific issue types (e.g., Bug, Feature, Task)
LabelsMatch issues with specific labels
AssigneeMatch issues assigned to specific users
StatusMatch issues in specific statuses (e.g., Ready, In Progress)
You can create multiple rulesets with different combinations of conditions. A plan is triggered when any ruleset matches.
Specifying a repository using Repository selector ensures plans are always generated against the correct codebase, skipping the repository resolution step.

ā€‹
Repository Resolution

Since Jira issues arenā€™t tied to a specific repository, CodeRabbit needs to determine which repository to analyze when generating a plan. CodeRabbit attempts to resolve the repository in this order:
1

Check Issue Description

If the issue description contains repository: <repo-name> or a link to a repository, CodeRabbit uses that repository.
repository: my-org/my-repo
OR
repository: https://github.com/my-org/my-repo
2

Check Planning Rulesets

If a planning ruleset specifies a repository, CodeRabbit uses that repository.
3

Automatic Detection

CodeRabbit leverages its deep understanding of your project to intelligently predict which repository is most relevant.
4

Ask the User

If CodeRabbit canā€™t confidently determine the repository, it replies to the issue asking you to specify which repository to use. Reply with the repository name to continue planning.

ā€‹
Viewing and Refining Plans

Once a Coding Plan is generated, view it in the CodeRabbit web app. Anyone in your organization can view and work with the plans.
Plans in the web app are organization-specific. If a plan shows a 404 error, verify youā€™re viewing the correct organization in the CodeRabbit web app.
View Coding Plans in the CodeRabbit web app

ā€‹
Chatting about Your Plan

Use the chat panel on the right side of the plan viewer in the CodeRabbit web app to:
  • Ask questions about the plan or the codebase
  • Request changes to specific tasks or phases
  • Challenge design choices and provide additional context
  • Get clarification on implementation details

ā€‹
Re-planning

After providing feedback through chat:
  1. Review your feedback in the chat history
  2. Click the Redo button
  3. CodeRabbit generates a new plan version incorporating your feedback

ā€‹
Version History

Each re-plan creates a new version. Use the version selector at the top of the plan viewer to:
  • View previous versions
  • Compare what changed between versions
  • Revert to an earlier version by marking it as active
You can only chat about the currently active version.

ā€‹
Handing Off to a Coding Agent

Click the Handoff button at the bottom of the plan viewer to see your options:

ā€‹
Copy to Clipboard

Copy the agentic prompts to your clipboard, then paste them into your preferred coding agent (Claude Code, Cursor, GitHub Copilot, etc.).

ā€‹
IDE Extension

If you have the CodeRabbit IDE extension installed, the Coding Plan can be sent directly to your coding agent through the extension. The agentic prompts appear in your coding agentā€™s input field, ready to execute. If you donā€™t have the extension installed, youā€™ll be prompted to install it.