Use Case: Bug Fixing

Intake format

Start with a reproducible report:

Symptom:
Environment:
Steps to reproduce:
Expected vs actual:
Logs/trace:
Regression window (if known):

Severity triage

Classify early to choose response speed:

P0: data loss, auth bypass, production outage
P1: major user flow broken
P2: degraded behavior with workaround
P3: minor/non-blocking

P0/P1 should always involve QA/security review.

Execution loop

Reproduce exactly in a minimal environment.
Isolate root cause (not just symptom patching).
Implement smallest safe fix.
Add regression tests for the failing path.
Re-check adjacent paths likely to share the same failure mode.

Prompt template for debug-agent

Bug: <error/symptom>
Repro steps: <steps>
Scope: <files/modules>
Expected behavior: <expected>
Need:
1) root cause
2) minimal fix
3) regression tests
4) adjacent-risk scan

Common escalation signals

Escalate to QA or security when bug touches:

authentication/session/token refresh
permission boundaries
payment/transaction consistency
performance regressions under load

Post-fix validation

original repro no longer fails
no new errors in related flows
tests fail before fix and pass after fix
rollback path is clear if hotfix is required

Done criteria

Bug-fixing is done when:

root cause is identified and documented
fix is verified through reproducible checks
regression coverage is in place