GitHub Leaked Secret Response Review (Defensive)
Review how the team handles an exposed credential.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
Review our procedure for a credential committed to a repository. Our current steps: [CURRENT_STEPS]. The type of credential (API key, token, password), by type only: [CREDENTIAL_TYPE]. Where the repository is visible (private, public): [VISIBILITY]. Tools we use for secret scanning: [SCANNING_TOOLS]. Compare each step with the order of priority: contain, rotate, assess, clean, prevent, learn. Mark steps in the wrong order, such as cleaning history before rotating; removing a secret from history does not make it safe. Write a one-page runbook as a table with step, owner, time target and evidence: 1 revoke or rotate immediately (first action within 30 minutes); 2 check logs for use between exposure and revocation; 3 assess what the credential could access and whether data may have been exposed, informing the security lead; 4 remove the secret from the code and consider history cleanup after rotation; 5 add prevention and record lessons. List prevention in order of cost: secret scanning, pre-commit checks, short-lived credentials, and a managed secrets store [VERIFY features]. Plan a quarterly drill in which a harmless dummy credential is committed to a test repository, the alert is raised, and the team practises and times the runbook. Keep the tone blame-free. Do not ask me to paste any real secret.
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.