#140Case study3 min
From detection to fix: the Hackaiz journey
Follow evidence, review a proposed change, approve its draft pull request and verify the deployed result.
Hackaiz team · updated 21/11/2026

02Read the finding before requesting a change
Use Danger, To fix and To watch to prioritize the findings, then read the evidence rather than acting on the severity alone. Establish which part of the system is involved, what was observed and what coverage the analysis actually obtained. A code finding and a live configuration finding may call for different responses. For example, changing source code does not itself demonstrate that the deployed service has picked up the change.
The Fixes page explains the starting point when no proposals exist: open a finding and ask Hackaiz to prepare one. This preparation is not permission to merge code or modify a customer database. Keep the original finding available during review so you can compare the proposed change with the observed problem. If the issue needs an external configuration change that cannot be automated, the documented path is guidance with targeted retesting, not a direct database write.
03Review the prepared proposal before approval
On Fixes, inspect Files concerned, What the change does, Reason for the change, Validation performed and Known limitations. Read them as separate questions: which files will change, how the change addresses the finding, what checks have been performed and what still needs human attention. Ready for review means your approval is needed; it does not mean the proposal has been merged, deployed or proven correct in every environment.
Only after reviewing the proposal should you use Apply this fix. The explicit approval authorizes delivery through a branch and a draft GitHub pull request. Hackaiz never merges that pull request automatically. A person must review the repository change and decide what to merge and deploy. If the attempt fails, open the finding to inspect the blocker instead of assuming the issue has been resolved merely because a proposal was prepared.
04Verify the changed system, not just the proposal
Use Check it again after the relevant change is in place, and read the resulting evidence. Repository tests and deployment evidence contribute to validation of code corrections. A draft pull request existing on GitHub is not proof that production has changed. If you altered an access rule or another external setting, verify the affected target rather than relying on a successful source-code test that never exercised that setting.
The project page offers Scan history and comparison with the previous analysis. Use the relevant scans to understand whether the previously observed evidence remains, while checking that their targets and coverage make the comparison meaningful. Keep review, merge and deployment responsibilities explicit in your team. Optional monitoring is available only for connected systems and consumes credits; it is a follow-up choice, not a requirement for reading a report or a guarantee that future changes cannot introduce risk.
Frequently asked questions
Take the next step
Was this article helpful?