# PBFuzz: Agentic Directed Fuzzing for PoV Generation

arXiv: https://arxiv.org/abs/2512.04611 · HTML: https://mlsanity.com/paper/2512.04611/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Agent threats, guardrails and governance, one of 16 directions in LLM agents and tool use.
- Builds on 26 of its 125 references that are in the map.
- First posted on arXiv 4 Dec 2025.

## Placed in

- LLM agents and tool use / [Agent threats, guardrails and governance](https://mlsanity.com/agents/agent-threats-guardrails-and-governance/)

## Builds on (most cited first, 26 of 26)

- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [Where LLM Agents Fail and How They can Learn From Failures](https://mlsanity.com/paper/2509.25370/) (2025)
- [OptimalThinkingBench: Evaluating Over and Underthinking in LLMs](https://mlsanity.com/paper/2508.13141/) (2025)
- [Can Large Language Models Write Good Property-Based Tests?](https://mlsanity.com/paper/2307.04346/) (2023)
- [Prompt Fuzzing for Fuzz Driver Generation](https://mlsanity.com/paper/2312.17677/) (2023)
- [KNighter: Transforming Static Analysis with LLM-Synthesized Checkers](https://mlsanity.com/paper/2503.09002/) (2025)
- [A Survey on Agentic Security: Applications, Threats and Defenses](https://mlsanity.com/paper/2510.06445/) (2025)
- [Angora: Efficient Fuzzing by Principled Search](https://mlsanity.com/paper/1803.01307/) (2018)
- [ELFuzz: Efficient Input Generation via LLM-driven Synthesis Over Fuzzer Space](https://mlsanity.com/paper/2506.10323/) (2025)
- [LLAMAFUZZ: Large Language Model Enhanced Greybox Fuzzing](https://mlsanity.com/paper/2406.07714/) (2024)
- [Locus: Agentic Predicate Synthesis for Directed Fuzzing](https://mlsanity.com/paper/2508.21302/) (2025)
- [ATLANTIS: AI-driven Threat Localization, Analysis, and Triage Intelligence System](https://mlsanity.com/paper/2509.14589/) (2025)
- [Agentic Property-Based Testing: Finding Bugs Across the Python Ecosystem](https://mlsanity.com/paper/2510.09907/) (2025)
- [PoCo: Agentic Proof-of-Concept Exploit Generation for Smart Contracts](https://mlsanity.com/paper/2511.02780/) (2025)
- [Cottontail: Large Language Model-Driven Concolic Execution for Highly Structured Test Input Generation](https://mlsanity.com/paper/2504.17542/) (2025)
- [UNIFUZZ: A Holistic and Pragmatic Metrics-Driven Platform for Evaluating Fuzzers](https://mlsanity.com/paper/2010.01785/) (2020)
- [MC2: Rigorous and Efficient Directed Greybox Fuzzing](https://mlsanity.com/paper/2208.14530/) (2022)
- [ODDFUZZ: Discovering Java Deserialization Vulnerabilities via Structure-Aware Directed Greybox Fuzzing](https://mlsanity.com/paper/2304.04233/) (2023)
- [HarnessAgent: Scaling Automatic Fuzzing Harness Construction with Tool-Augmented LLM Pipelines](https://mlsanity.com/paper/2512.03420/) (2025)
- [SAVIOR: Towards Bug-Driven Hybrid Testing](https://mlsanity.com/paper/1906.07327/) (2019)
- [autofz: Automated Fuzzer Composition at Runtime](https://mlsanity.com/paper/2302.12879/) (2023)
- [Large Language Model Powered Symbolic Execution](https://mlsanity.com/paper/2505.13452/) (2025)
- [Fuzzing: Randomness? Reasoning! Efficient Directed Fuzzing via Large Language Models](https://mlsanity.com/paper/2507.22065/) (2025)
- [The Progress, Challenges, and Perspectives of Directed Greybox Fuzzing](https://mlsanity.com/paper/2005.11907/) (2020)
- [HyperGo: Probability-based Directed Hybrid Fuzzing](https://mlsanity.com/paper/2307.07815/) (2023)

## Built on by (most cited first, 4 of 4)

- [Patch-to-PoC: A Systematic Study of Agentic LLM Systems for Linux Kernel N-Day Reproduction](https://mlsanity.com/paper/2602.07287/) (2026)
- [Execution-State-Aware LLM Reasoning for Automated Proof-of-Vulnerability Generation](https://mlsanity.com/paper/2602.13574/) (2026)
- [To Defend Against Cyber Attacks, We Must Teach AI Agents to Hack](https://mlsanity.com/paper/2602.02595/) (2026)
- [SyzHarness: Patch-Based Kernel Bug Reproduction with LLM-Synthesized Fuzzing Harnesses](https://mlsanity.com/paper/2609.23889/) (2026)

## Abstract (arXiv)

Proof-of-Vulnerability (PoV) input generation is a critical task in software security and supports downstream applications such as path generation and validation. Generating a PoV input requires solving two sets of constraints: (1) reachability constraints for reaching vulnerable code locations, and (2) triggering constraints for activating the target vulnerability. Existing approaches, including directed greybox fuzzing and LLM-assisted fuzzing, struggle to efficiently satisfy these constraints. This work presents an agentic method that mimics human experts. Human analysts iteratively study code to extract semantic reachability and triggering constraints, form hypotheses about PoV triggering strategies, encode them as test inputs, and refine their understanding using debugging feedback. We automate this process with an agentic directed fuzzing framework called PBFuzz. PBFuzz tackles four challenges in agentic PoV generation: autonomous code reasoning for semantic constraint extraction, custom program-analysis tools for targeted inference, persistent memory to avoid hypothesis drift, and property-based testing for efficient constraint solving while preserving input structure. Experiments on the Magma benchmark show strong results. PBFuzz triggered 57 vulnerabilities, surpassing all baselines, and uniquely triggered 17 vulnerabilities not exposed by existing fuzzers. PBFuzz achieved this within a 30-minute budget per target, while conventional approaches use 24 hours. Median time-to-exposure was 339 seconds for PBFuzz versus 8680 seconds for AFL++ with CmpLog, giving a 25.6x efficiency improvement with an API cost of 1.83 USD per vulnerability. In real-world application, PBFuzz reproduced three FFmpeg 1-day CVEs that had no public PoVs.