fuzzing-dictionary

A fuzzing dictionary improves parser testing by providing domain-specific tokens that guide fuzzers toward meaningful edge cases, accelerating the discovery of vulnerabilities in file format decoders and protocols.

Create a text file containing quoted strings, with one token per line, and save it as a .dict file. Integrate this fuzzing dictionary into libFuzzer by running your fuzz target with the -dict flag pointing to the file.

Yes, a fuzzing dictionary works with AFL++ and cargo-fuzz workflows. You can supply the dictionary file using the -x flag for AFL++ and the -dict flag for cargo-fuzz to guide input mutations.

Fuzzers often find only trivial inputs without a dictionary because they lack domain-specific tokens like magic numbers and keywords, preventing them from reaching deeper logic paths in complex parsers and protocols.

You need a fuzzing dictionary for security testing when targeting parsers, protocols, or file format decoders, as it provides structured tokens that steer input mutation toward meaningful edge cases and vulnerabilities.