GitHub
Add codebase context to PRDs
The GitHub integration gives Flowback's AI context about your codebase, resulting in more specific and actionable PRDs. When GitHub is connected, the AI can reference relevant files, recent commits, and repository structure when generating product requirement documents.
Connecting GitHub
- Navigate to Integrations in your dashboard
- Click Connect GitHub
- Authorize Flowback on GitHub (requires
repoandread:orgscopes) - Select the repository you want to index
Codebase indexing
Once connected, Flowback builds a codebase index that includes:
- File tree — The complete directory structure of your repository
- README — Your project's README for understanding the product
- Recent commits — The latest commit history for context on what's changed
This index is used by the AI during PRD generation to reference specific parts of your codebase that may be relevant to the feedback.
How it improves PRDs
With GitHub connected, PRDs include additional context:
- Relevant files — The AI identifies which source files are likely related to the feedback
- Recent changes — If a bug might be related to a recent commit, the AI notes this
- Technical context — The AI can reference your tech stack and project structure in its analysis
This extra context helps your engineering team start investigating issues faster, as they already have pointers to the relevant parts of the codebase.
Privacy and security
Flowback uses the GitHub OAuth token to read your repository data. The token is stored encrypted and is only used to build the codebase index. Your source code is not stored — only metadata (file paths, commit messages, README content) is retained.
You can disconnect GitHub at any time to revoke access. Disconnecting removes the stored token and codebase index.