agentsastLast reviewed 2026-09-13

Proof-of-concept harness

Direct answerTooling that turns a candidate finding into an executable exploit or failing test, such as zkao's cheatVM for Aleo or Codex Security's sandbox.

In more detail

A finding with a runnable PoC costs minutes to validate; one without can cost hours. The HumanityLink engagement found AI-written PoCs unrealistic without such tooling.

Tools that address it

zkao, Olympix, Codex Security (formerly Aardvark).

False positive rate, Precision vs recall, Agentic scanning, LLM plus fuzzing, LLM plus symbolic execution or formal verification, Hallucinated vulnerabilities, Triage burden, Benchmark contamination, Human-in-the-loop, AI-assisted audit vs AI audit, Prompt injection in auditing pipelines, Responsible disclosure of AI-found bugs, Continuous scanning and run-count coverage, Threat model file, Severity calibration

Getting help

Firms on this index that handle this in practice: zkSecurity, Trail of Bits, Zellic, Nethermind Security.