Goal
Use the prebuilt PawSQL repository connector to review SQL on GitHub pushes and pull requests.Before you begin
- Verify connectivity from PawSQL to GitHub or your GitHub Enterprise instance.
- Prepare a GitHub identity or credential with access to the target repositories.
- Create the PawSQL project, workspace, and review policy.
- Identify the organizations, repositories, branches, and SQL paths in scope.
- Define pull-request gate and feedback behavior.
Configuration
1
Create the GitHub connection
Choose GitHub in the PawSQL repository connectors and complete endpoint and identity configuration.
2
Select organizations and repositories
Authorize only the repositories that PawSQL must review.
3
Choose trigger events
Enable the required push, pull-request creation, synchronization, or reopen events.
4
Map review context
Set target branches, SQL paths, extensions, workspace, and review policy.
5
Configure the gate
Define blocking severity, failure behavior, check status, and report feedback.
6
Run acceptance tests
Open a test pull request and exercise pass, block, no-SQL, and duplicate-event cases.
Feedback
The result should reference the current revision and expose the gate decision, finding count, and PawSQL report link. When a pull request changes, a result for an older revision must not satisfy the merge requirement.Permission guidance
- Grant only the repository-read and check-status permissions required.
- Limit organization-level authorization to approved repositories.
- Rotate credentials and review stale authorizations.
- Keep tokens out of logs and exported configuration.
- Handle pull requests from forks without exposing privileged credentials.
Expected Result
A review job is created for relevant SQL changes, and the result references the current revision with the gate decision, finding count, and report link.The prebuilt GitHub repository connector is distinct from a GitHub Actions OpenAPI workflow. If the pipeline must control the review stage directly, use the OpenAPI pipeline pattern.