# Projects and products consuming, wrapping, and using garak

## Tooling 

(open-source / free only)



| Name | Licence | Description |
| - | - | - |
| [TrustyAI Garak](https://github.com/trustyai-explainability/llama-stack-provider-trustyai-garak) | Apache |  Out-of-Tree Llama Stack Eval Provider for Garak Red Teaming |
| [Garak-MCP](https://github.com/EdenYavin/Garak-MCP) | MIT license | MCP Server for garak |
| [Garak Report](https://github.com/lreading/garak-repo) | Apache-2.0 | A repository for your Garak runs, as well as a modern visualizer. |

<!-- * [Garak Studio](https://github.com/hasbegun/garak_studio) - OSS GUI for garak -->


## Integrations

| Name | Description |
| - | - |
| [Tumeryk](http://www.tumeryk.com) | The Tumeryk platform empowers organizations to safeguard AI systems, ensuring secure, reliable, and policy-aligned deployments. Scan LLMs & endpoints to prevent jailbreaks, data leaks, and IP exposure |
| [Vijil](https://www.vijil.ai/) | Vijil helps organizations build and operate autonomous agents that humans can trust. Garak forms part of their “vijil score”. They have an API and efficient orchestration engine |
| [Deepchecks](https://deepchecks.com) | [Integrating garak and NeMo Guardrails together](https://deepchecks.com/the-best-llm-safety-net-to-date-deepchecks-garak-and-nemo-guardrails-all-in-one-bundle/); [Example use of garak tools](https://llmdocs.deepchecks.com/docs/pentesting-your-llm-pipeline) |
| [Mindgard](https://www.mindgard.ai) | An AI security platform, Mindgard integrate garak as part of their pentesting & eval suite |
| [Giskard](https://giskard.ai) | [Giskard integration](https://docs.giskard.ai/en/stable/reference/scan/llm_detectors.html) |
| [OpsMX](https://www.opsmx.com/) | "OpsMx Delivery Shield embeds Garak’s adversarial testing engine to continuously probe, monitor, and guard live AI and LLM workloads against jailbreaks, data leaks, and policy violations"  https://www.opsmx.com/dynamic-runtime-ai-security |
| [Upwind](https://www.upwind.io/) | "Upwind secures your cloud deployments, configurations, and applications through a runtime fabric that provides real-time visibility from the inside out" |
| [GuardionAI](https://guardion.ai/) | "Discover, analyze, and understand 89 AI security threats across 25 categories. From prompt injection to data leakage, including 29,310 prompt attack examples. Explore the complete landscape of LLM vulnerabilities." [PromptIntel](https://guardion.ai/promptintel) |


## Want to be added?
Want a project mentioned, or the description updated? Send a pull request with details. Please use a [neutral tone](https://en.wikipedia.org/wiki/Wikipedia:Neutral_point_of_view).

Inclusion on this page does not imply an endorsement by the maintainers or NVIDIA.