Skip to content
Latchkey

Fuzz Test - CI/CD Glossary Definition

A fuzz test feeds a program large volumes of random, malformed, or unexpected input to find crashes, hangs, and security bugs that example-based tests miss.

Coverage-guided fuzzing

Modern fuzzers like libFuzzer and AFL watch code coverage and mutate inputs to reach new branches, steadily exploring more of the program. Googles OSS-Fuzz runs this continuously on open-source projects.

In CI

A short fuzzing budget per pull request catches new crash regressions, while long-running fuzzers run out-of-band because exhaustive fuzzing is too slow for the critical path.

Related guides

Tired of flaky CI? Latchkey auto-heals failed jobs and retries them for you. Start free →