Smoke Test
A software testing method that quickly verifies whether an application's core paths are functional after each build, so that serious problems are found as early as possible. It is low in cost and fast to execute.
Core Facts
Timeline (last 90 days)
What smoke testing and regression verification have in common is that their judgment criteria are relatively clear (a certain element appears in the interface, the flow reaches a certain step), making them suitable for AI agents to execute; whereas experience issues, boundary conditions and exception paths that require subjective judgment are still better suited to manual acceptance.
If a case never fails on any backend tried, it is a smoke test rather than a regression test and should be placed in a separate file.
The smoke test works by hitting the routing endpoint with 20 concurrent requests and tallying HTTP status codes and response times, in order to catch rate limiting or cold-start issues on free instances on deployment day.
A layered testing strategy divides tests into fast smoke tests and full evaluation tests, running the full suite only before merging, which can reduce unnecessary model calls.
When deploying local AI models, a smoke test using a short audio clip of approximately 30 seconds should be performed first to verify at low cost whether the environment and model are compatible, avoiding crashes after a long generation process
All Facts (6)
Smoke Testing only checks whether the system can start and whether core paths can run through, while Regression Testing is a comprehensive validation of all existing functionality, and the two cannot substitute for each other.
65%UnverifiedWhat smoke testing and regression verification have in common is that their judgment criteria are relatively clear (a certain element appears in the interface, the flow reaches a certain step), making them suitable for AI agents to execute; whereas experience issues, boundary conditions and exception paths that require subjective judgment are still better suited to manual acceptance.
50%UnverifiedIf a case never fails on any backend tried, it is a smoke test rather than a regression test and should be placed in a separate file.
50%UnverifiedThe smoke test works by hitting the routing endpoint with 20 concurrent requests and tallying HTTP status codes and response times, in order to catch rate limiting or cold-start issues on free instances on deployment day.
50%UnverifiedA layered testing strategy divides tests into fast smoke tests and full evaluation tests, running the full suite only before merging, which can reduce unnecessary model calls.
50%UnverifiedWhen deploying local AI models, a smoke test using a short audio clip of approximately 30 seconds should be performed first to verify at low cost whether the environment and model are compatible, avoiding crashes after a long generation process
50%