| name | exploratory-testing |
| description | Execute a focused exploratory test session that probes real user behavior, edge cases, state transitions, and unexpected interactions. |
Exploratory Testing
Use when behavior needs human-style probing beyond scripted assertions.
Procedure
- Define a short test charter: feature, risk, user goal, and timebox.
- Establish the expected happy path and known constraints.
- Explore boundaries: empty/large/invalid inputs, retries, interruption, navigation changes, persistence, permissions, concurrency, slow/failing dependencies, and device/layout differences where relevant.
- Vary sequence and state rather than repeating the same path.
- Capture defects immediately with exact reproduction steps, environment, evidence, expected result, and actual result.
- Note suspicious behavior that needs follow-up separately from confirmed bugs.
- Summarize coverage, findings, and remaining uncertainty at session end.
Quality gate
A defect is not useful until another person can reproduce it. A clean session is not proof of absence; state what was and was not explored.
