Short answer
Debugging assesses structured investigation from an observed symptom to a supported cause and verified correction.
About Debugging
The ability to find the causes of unexpected behavior through structured investigation, controlled experiments, and relevant evidence.
Use this competency for
- Roles that investigate defects, incidents, data discrepancies, or unexpected technical behavior.
- Functions where unsupported fixes can hide or repeat failures.
Do not use this competency for
- Roles that report issues but do not investigate technical causes or verify fixes.
Important distinctions
Testing strategy
Testing strategy decides how to expose failures. Debugging investigates a failure that has already been observed.
Reliability engineering
Reliability engineering improves system availability and recovery. Debugging isolates causes within a specific failure investigation.
Expectations by level
IC1
Individual contributor 1
Investigates a bounded problem with guidance, reproduces the symptom, and records evidence that separates facts from assumptions.
Observable behaviors
- Captures the inputs and environment needed to reproduce the issue.
- Changes one relevant variable at a time during investigation.
- Verifies the fix against the original symptom.
Examples
- Reduced a failing request to one input that reproduced the error locally.
- Compared logs before and after a fix to confirm the failing branch no longer ran.
IC2
Individual contributor 2
Independently isolates causes in unfamiliar team-owned systems, selecting useful evidence and testing competing explanations.
Observable behaviors
- Builds a timeline from logs, metrics, traces, or data changes.
- Ranks hypotheses and states what evidence would reject each one.
- Checks for related failure paths before closing the investigation.
Examples
- Used a trace to separate network delay from slow database work.
- Found that a data race, not the visible validation error, caused intermittent failures.
IC3
Individual contributor 3
Leads ambiguous investigations across systems and teams, establishes shared diagnostic methods, and reduces future time to isolate similar failures.
Observable behaviors
- Coordinates evidence collection across owned system boundaries.
- Identifies missing diagnostic signals revealed by an investigation.
- Turns a recurring investigation path into a reusable tool or guide.
Examples
- Aligned timestamps from three systems to locate a failure at a handoff boundary.
- Added correlation data and an investigation guide after repeated incidents lacked a common trace.