Skip to main content
PawSQL Client for DBeaver inserts a review step into the SQL execution path. Before DBeaver sends a statement to the database, the extension requests a static PawSQL review and uses the configured mode and severity threshold to decide whether to proceed, request confirmation, or stop.

When to use it

  • Catch UPDATE or DELETE statements that lack required safeguards.
  • Surface operational risk before running DDL.
  • Detect scans, ineffective predicates, and risky joins.
  • Apply a consistent pre-execution check to production datasources.
  • Give developers immediate feedback on team SQL standards.

Prerequisites

  • A compatible DBeaver installation.
  • The official PawSQL DBeaver package or installation endpoint.
  • Network access to PawSQL Cloud or PawSQL Server.
  • Permission to use the intended project, workspace, and review policy.
  • A reliable classification of development, test, and production datasources in DBeaver.

Install the extension

Install PawSQL for DBeaver from a software repository; see Install IDE Extensions for steps. After installation, confirm that PawSQL Client settings appear under Preferences.
Installation entry points and compatible versions can change with DBeaver and PawSQL releases. Follow the instructions shipped with the extension package you are installing.

Connect to PawSQL

Under Preferences > PawSQL Client, configure the endpoint and authentication required by your deployment. Select or verify the project, workspace, and review policy used for analysis. Test the integration with a read-only statement in a non-production environment before enabling enforcement on production datasources.

PawSQL Client settings in Preferences

Choose a review mode

Start with Prompt mode on test datasources to validate policy behavior and latency. Move critical datasources to Enforce mode after the policy and exclusions are reviewed.

Set the severity threshold

Review findings use the following levels: The threshold determines which levels require confirmation. With Warning as the threshold, both Critical and Warning findings require confirmation, while Info findings can proceed.
After changing the threshold, validate the effective behavior with representative SQL. Use the configuration page’s displayed interception scope instead of inferring strictness from the label alone.

Exclude trusted datasources

Datasource exclusions are intended for environments where pre-execution review is deliberately unnecessary, such as an isolated local database. An excluded datasource follows DBeaver’s normal execution path regardless of the selected review mode. Before adding an exclusion:
  • make sure the name and connection identity uniquely identify the datasource;
  • avoid broad naming or matching patterns;
  • verify that copied production and staging connections are not excluded accidentally;
  • assign an owner and review the exclusion list regularly.

Execute and review

1

Run the statement

Use the normal execution action in the DBeaver SQL editor.
2

Wait for static review

The extension sends the SQL and required context to PawSQL. This check does not execute the statement or create a candidate index.
3

Inspect the findings

Review the highest severity, triggered rules, affected fragments, and any available rewrite or index guidance.
4

Make an execution decision

Revise and review the SQL again, cancel the operation, or proceed after confirming that the residual risk is acceptable.
Risk prompt from the pre-execution review

Risk prompt from the pre-execution review

Proceeding acknowledges the current prompt; it is not a substitute for formal change approval in a controlled environment.

Expected Result

With review enabled, the extension returns rule findings with severity and, when the configured threshold is reached, shows a risk prompt where you can revise the SQL, cancel, or proceed.

Verification

Confirm the integration by running a read-only statement on a non-production datasource and checking that expected findings or the risk prompt appear, and that risky SQL is intercepted according to the configured threshold.

Pre-execution review versus deep optimization

Pre-execution review is intentionally low-latency and side-effect free. Use full optimization when you need plan comparison, index-benefit analysis, or runtime measurements.

Troubleshooting

Rollout guidance

  1. Validate connectivity and findings against development or test datasources.
  2. Run Prompt mode with representative SQL and collect feedback.
  3. Refine the policy, severity threshold, and exclusions.
  4. Enable Enforce mode on selected critical datasources.
  5. Define bypass, outage fallback, and audit-record requirements.
  6. Review exclusions and high-risk dispositions regularly.

SQL Review

Review Policy

Read Review Results

Control High-Risk SQL