0sec Labs Releases Open-Source AI Security Harness
Zurich-based 0sec Labs has released 0, an open-source harness for AI-assisted vulnerability investigations, code review, penetration testing, reverse engineering and fix generation. The software is distributed under the MIT OR Apache-2.0 license as a research preview. It is a control layer rather than an AI model: users connect their own provider, with 14 providers listed, while inference runs externally.
The system can be operated through a local browser console, a headless command-line interface or an MCP server for coding agents. It supports Apple Silicon macOS, x64 and ARM64 Linux, as well as the ghcr.io/0sec-labs/0 container, and does not require a GPU. Its workflows plan and investigate targets, filter findings, reproduce candidates in a separate session and generate HTML, PDF, SARIF, Markdown or JSON reports. Compiled targets can be analyzed through 0verse, which integrates Ghidra, angr and AFL++.
The company reports, with results described as self-reported, a top score of 93/95 on XBOW web CTFs and 81/104 in a black-box XBOW union. It also says the project has contributed to more than 30 merged Linux kernel zero-day fixes. A hosted edition, 0cloud, is currently waitlisted, while the open-source version is free apart from model-provider charges.
Written by Sakhanda Wire AI from reporting by MarkTechPost. Facts, quotes and full context are in the original article below.