Strix
An AI penetration testing tool that runs locally with Docker, supports Ollama and LM Studio or cloud LLMs, and uses the Apache 2.0 license.
Strix is an AI penetration testing tool for developers and security teams assessing code, web applications and APIs. Its free, open source edition runs locally with Docker under the Apache 2.0 license. It supports local models through Ollama or LM Studio, as well as remote providers such as OpenAI and Anthropic; cloud-backed runs send model requests to the selected provider.
Agents test suspected vulnerabilities with proof-of-concept exploits. They combine source analysis with tests against running applications, using browser automation and HTTP request manipulation through Caido. Tests cover access control, injection, authentication and business logic flaws, including workflow bypasses and race conditions.
Agents share their discoveries. They can coordinate reconnaissance and exploitation across targets and combine vulnerabilities into attack paths. Findings include exploit evidence and remediation guidance, and Strix can generate security patches and penetration testing reports. API testing can use OpenAPI, Swagger or Postman definitions to identify endpoints.
Scan results stay on disk, with a local web dashboard for findings, agent activity and past runs. GitHub Actions and other CI pipelines can run checks on pull requests, and coding assistants including Claude Code, Cursor and Codex can use Strix through agent skills.
Strix Cloud is a separate managed service that can receive uploaded source code and run scans remotely. It offers continuous testing and security patches as pull requests, with GitHub and GitLab integrations.
