[{"data":1,"prerenderedAt":4},["ShallowReactive",2],{"readme:deepteam":3},"\u003Cp align=\"center\">\n    \u003Cpicture>\n        \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"assets\u002Fhero\u002Fwordmark-dark-v2.svg\">\u003C\u002Fsource>\n        \u003Cimg alt=\"DeepTeam.\" src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Fconfident-ai\u002Fdeepteam\u002FHEAD\u002Fassets\u002Fhero\u002Fwordmark-light-v2.svg\" width=\"520\" \u002F>\n    \u003C\u002Fpicture>\n\u003C\u002Fp>\u003Ch1>The LLM Red Teaming Framework\u003C\u002Fh1>\u003Ch4>\n    \u003Cp>\n        \u003Ca href=\"https:\u002F\u002Fwww.trydeepteam.com?utm_source=GitHub\" rel=\"nofollow ugc noopener\">Documentation\u003C\u002Fa> |\n        \u003Ca href=\"#-vulnerabilities-attacks-and-features\" rel=\"nofollow ugc noopener\">Vulnerabilities, Attacks, and Features\u003C\u002Fa> |\n        \u003Ca href=\"#-quickstart\" rel=\"nofollow ugc noopener\">Getting Started\u003C\u002Fa> |\n        \u003Ca href=\"#deepteam-with-confident-ai\" rel=\"nofollow ugc noopener\">Confident AI\u003C\u002Fa>\n    \u003C\u002Fp>\u003Cp>\n\u003C\u002Fp>\u003C\u002Fh4>\u003Cp align=\"center\">\n    \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fconfident-ai\u002Fdeepteam\u002Freleases\" rel=\"nofollow ugc noopener\">\n        \u003Cimg alt=\"GitHub release\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fv\u002Frelease\u002Fconfident-ai\u002Fdeepteam\" \u002F>\n    \u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fdiscord.gg\u002F3SEyvpgu2f\" rel=\"nofollow ugc noopener\">\n        \u003Cimg alt=\"discord-invite\" src=\"https:\u002F\u002Fdcbadge.limes.pink\u002Fapi\u002Fserver\u002F3SEyvpgu2f?style=flat\" \u002F>\n    \u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fconfident-ai\u002Fdeepteam\u002Fblob\u002Fmain\u002FLICENSE.md\" rel=\"nofollow ugc noopener\">\n        \u003Cimg alt=\"License\" src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flicense\u002Fconfident-ai\u002Fdeepteam.svg?color=yellow\" \u002F>\n    \u003C\u002Fa>\n\u003C\u002Fp>\u003Cp align=\"center\">\n    \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Fconfident-ai\u002Fdeepteam?lang=de\" rel=\"nofollow ugc noopener\">Deutsch\u003C\u002Fa> | \n    \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Fconfident-ai\u002Fdeepteam?lang=es\" rel=\"nofollow ugc noopener\">Español\u003C\u002Fa> | \n    \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Fconfident-ai\u002Fdeepteam?lang=fr\" rel=\"nofollow ugc noopener\">français\u003C\u002Fa> | \n    \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Fconfident-ai\u002Fdeepteam?lang=ja\" rel=\"nofollow ugc noopener\">日本語\u003C\u002Fa> | \n    \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Fconfident-ai\u002Fdeepteam?lang=ko\" rel=\"nofollow ugc noopener\">한국어\u003C\u002Fa> | \n    \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Fconfident-ai\u002Fdeepteam?lang=pt\" rel=\"nofollow ugc noopener\">Português\u003C\u002Fa> | \n    \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Fconfident-ai\u002Fdeepteam?lang=ru\" rel=\"nofollow ugc noopener\">Русский\u003C\u002Fa> | \n    \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Fconfident-ai\u002Fdeepteam?lang=zh\" rel=\"nofollow ugc noopener\">中文\u003C\u002Fa>\n\u003C\u002Fp>\u003Cp>\u003Cstrong>DeepTeam\u003C\u002Fstrong> is a simple-to-use, open-source red teaming framework for LLM systems. Think of it as penetration testing, but for LLMs.\u003C\u002Fp>\n\u003Cp>DeepTeam simulates attacks — jailbreaking, prompt injection, multi-turn exploitation, and more — to uncover vulnerabilities like bias, PII leakage, and SQL injection in your AI agents, RAG pipelines, and chatbots. It also offers \u003Cstrong>guardrails\u003C\u002Fstrong> to prevent these issues in production.\u003C\u002Fp>\n\u003Cp>DeepTeam runs \u003Cstrong>locally on your machine\u003C\u002Fstrong> and is built on \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fconfident-ai\u002Fdeepeval\" rel=\"nofollow ugc noopener\">DeepEval\u003C\u002Fa>, the open-source LLM evaluation framework.\u003C\u002Fp>\n\u003Cblockquote>\n\u003Cp>[!IMPORTANT]\nNeed a place for your red teaming results to live? Sign up to the \u003Ca href=\"https:\u002F\u002Fapp.confident-ai.com?utm_source=GitHub\" rel=\"nofollow ugc noopener\">Confident AI\u003C\u002Fa> platform to manage risk assessments, monitor vulnerabilities in production, and share reports with your team.\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Cp align=\"center\">\n    \u003Cimg src=\"https:\u002F\u002Fgithub.com\u002Fconfident-ai\u002Fdeepteam\u002Fblob\u002Fmain\u002Fassets\u002Fconfident-demo.gif\" alt=\"Confident AI + DeepTeam\" width=\"100%\" \u002F>\n\u003C\u002Fp>\u003Cblockquote>\n\u003Cp>Want to talk LLM security, need help picking attacks, or just to say hi? \u003Ca href=\"https:\u002F\u002Fdiscord.com\u002Finvite\u002F3SEyvpgu2f\" rel=\"nofollow ugc noopener\">Come join our discord.\u003C\u002Fa>\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Cp> \u003C\u002Fp>\n\u003Ch1>🔥 Vulnerabilities, Attacks, and Features\u003C\u002Fh1>\n\u003Cul>\n\u003Cli>\u003Cp>📐 50+ ready-to-use \u003Ca href=\"https:\u002F\u002Fwww.trydeepteam.com\u002Fdocs\u002Fred-teaming-vulnerabilities\" rel=\"nofollow ugc noopener\">vulnerabilities\u003C\u002Fa> (all with explanations) powered by \u003Cstrong>ANY\u003C\u002Fstrong> LLM of your choice. Each vulnerability uses LLM-as-a-Judge metrics that run \u003Cstrong>locally on your machine\u003C\u002Fstrong> to produce binary pass\u002Ffail scores with reasoning:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cdetails>\n\u003Csummary>\u003Cb>Data Privacy\u003C\u002Fb>\u003C\u002Fsummary>\u003Cul>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fwww.trydeepteam.com\u002Fdocs\u002Fred-teaming-vulnerabilities-pii-leakage\" rel=\"nofollow ugc noopener\">PII Leakage\u003C\u002Fa> — disclosure of sensitive personal information\u003C\u002Fli>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fwww.trydeepteam.com\u002Fdocs\u002Fred-teaming-vulnerabilities-prompt-leakage\" rel=\"nofollow ugc noopener\">Prompt Leakage\u003C\u002Fa> — exposure of system prompt secrets and instructions\u003C\u002Fli>\n\u003C\u002Ful>\n\u003C\u002Fdetails>\n\u003C\u002Fli>\n\u003Cli>\u003Cdetails>\n\u003Csummary>\u003Cb>Responsible AI\u003C\u002Fb>\u003C\u002Fsummary>\u003Cul>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fwww.trydeepteam.com\u002Fdocs\u002Fred-teaming-vulnerabilities-bias\" rel=\"nofollow ugc noopener\">Bias\u003C\u002Fa> — stereotypes and unfair treatment across gender, race, religion, politics\u003C\u002Fli>\n\u003Cli>[Toxicity](\u003Ca href=\"https:\u002F\u002Fwww.trydeep\" rel=\"nofollow ugc noopener\">https:\u002F\u002Fwww.trydeep\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003C\u002Fdetails>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003C\u002Fli>\n\u003C\u002Ful>\n",1784240407147]