# google/oss-fuzz-gen

LLM powered fuzzing via OSS-Fuzz.

Repository: https://github.com/google/oss-fuzz-gen
Canonical: https://ross.abutalabs.com/products/oss-fuzz-gen
Language: Python
License: Apache-2.0
License Family: permissive
Topics: ai, fuzzing, llm, security
Last push: 2026-03-17T19:15:40+00:00

## Health v2 (maintenance only)
Score: 58/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 72, release rhythm 35, longevity 68
- inputs: {"age_days": 952, "days_push": 169, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1431, forks 223 (observed 2026-08-28T04:04:42.879194+00:00)

## What it is
A Google framework that uses large language models to automatically generate fuzz targets for real-world C/C++, Java, and Python projects, benchmarking them via the OSS-Fuzz platform. It evaluates generated targets on compilability, crashes, and coverage, and has discovered real vulnerabilities in open-source projects.

## Use cases
- generate fuzz targets automatically with LLMs
- find vulnerabilities in open-source C/C++ projects
- benchmark LLM code generation for fuzzing
- evaluate coverage of generated fuzz harnesses
- run agent-based fuzz target experiments
- compare Gemini and GPT models on fuzzing tasks

## When to choose
- you maintain an OSS-Fuzz project and want AI-generated fuzz targets
- you're researching LLM effectiveness for security testing
- you want to scale fuzz target writing across many projects

## When to avoid
- you need a general-purpose fuzzer rather than target generation
- you can't access commercial LLM APIs
- your project isn't compatible with OSS-Fuzz

## Facets
- artifact type: framework
- maturity: active
- function: fuzzing, llm-inference, agent-framework, security, benchmarking, testing
- domain: security, developer-tools, large-language-models, testing, artificial-intelligence
- platform: python, cloud
- tags: fuzz-target-generation, oss-fuzz, llm-benchmarking, vulnerability-discovery, automated-testing, linux, docker

## Member repositories
- google/oss-fuzz-gen (main) score 58

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:42.879194+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:37:07.383996+00:00, confidence not recorded.
  - readme: https://github.com/google/oss-fuzz-gen (fetched 2026-08-28T04:04:42.879194+00:00, sha f48d657d3fd0)
- Data as of 2026-08-30T08:39:29.467469+00:00.