How to Teach Debugging with Playbooks Instead of Guesswork
Beginners often jump straight to random fixes. A debugging playbook teaches ordered reasoning and helps teams communicate clearly under pressure.
Step 1: Capture observable symptom first
Symptom: API timeout on checkout screen after 3rd retry.
Scope: iOS + Android
Frequency: 40% on mobile data
Step 2: Build hypothesis table before code changes
H1: DNS failover delay
H2: stale auth token refresh loop
H3: backend rate limit burst
Step 3: Run minimal experiments with clear stop criteria
./run_probe.sh --scenario=mobile-data --retries=5 --log-json
Pitfall
Editing multiple layers at once. You lose causal signal and cannot prove which change fixed it.
Verification
- Each bug report includes symptom, scope, hypothesis, and experiment result.
- Fixes map to validated root causes.
- Postmortems improve the playbook library.