0sec Labs Open Sources AI Security Harness with Web UI and CLI
Executive Briefing
Synthesized and verified with gemini-aistudio-api-key/models/gemini-3.5-flash-lite / reviewer requested: gemini/gemini-3.5-flash-lite
“Zurich-based startup 0sec Labs has released 0, an open-source AI security harness featuring...”
Key Takeaways
- 1.The open-source AI security harness named 0 was released by 0sec Labs under an MIT or Apache-2.0 license as a research preview.
- 2.The system offers multiple interfaces including a local browser console, a headless CLI, and an MCP server for coding agents.
- 3.Users must bring their own model from 14 listed providers or Claude, Codex, and Gemini CLI runtimes.
Zurich-based startup 0sec Labs has released 0, an open-source AI security harness featuring a local browser console, CLI, and MCP server. The tool coordinates vulnerability investigations, code reviews, and reverse engineering tasks using user-supplied large language models.
Key Developments and Technical Context
The open-source AI security harness named 0 was released by 0sec Labs under an MIT or Apache-2.0 license as a research preview.
The system offers multiple interfaces including a local browser console, a headless CLI, and an MCP server for coding agents.
Users must bring their own model from 14 listed providers or Claude, Codex, and Gemini CLI runtimes.
Strategic Analysis and Industry Implications
Combining a CLI, web console, and MCP server enables development teams to integrate automated security testing workflows directly into their daily environments.
The bring-your-own-model approach gives users the flexibility to choose providers that align with their organizational cost and data security requirements.
Outlook and Key Factors to Watch
The actual performance and accuracy of the security harness in large enterprise environments remain unverified independently beyond self-reported scores.
