[{"data":1,"prerenderedAt":4},["ShallowReactive",2],{"readme:giskard-oss":3},"\u003Cp align=\"center\">\n  \u003Cimg alt=\"giskardlogo\" src=\"https:\u002F\u002Fraw.githubusercontent.com\u002FGiskard-AI\u002Fgiskard-oss\u002FHEAD\u002Freadme\u002Flogo_light.png#gh-light-mode-only\" \u002F>\n  \u003Cimg alt=\"giskardlogo\" src=\"https:\u002F\u002Fraw.githubusercontent.com\u002FGiskard-AI\u002Fgiskard-oss\u002FHEAD\u002Freadme\u002Flogo_dark.png#gh-dark-mode-only\" \u002F>\n\u003C\u002Fp>\n\u003Ch1>Evals, Red Teaming and Test Generation for Agentic Systems\u003C\u002Fh1>\n\u003Ch3>Modular, Lightweight, Dynamic and Async-first \u003C\u002Fh3>\n\u003Cdiv align=\"center\">\u003Cp>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FGiskard-AI\u002Fgiskard\u002Freleases\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fv\u002Frelease\u002FGiskard-AI\u002Fgiskard\" alt=\"GitHub release\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FGiskard-AI\u002Fgiskard\u002Fblob\u002Fmain\u002FLICENSE\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-Apache_2.0-blue.svg\" alt=\"License\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fpepy.tech\u002Fproject\u002Fgiskard\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fstatic.pepy.tech\u002Fbadge\u002Fgiskard\u002Fmonth\" alt=\"Downloads\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FGiskard-AI\u002Fgiskard-oss\u002Factions\u002Fworkflows\u002Fci.yml\u002Fbadge.svg?branch=main\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fgithub.com\u002FGiskard-AI\u002Fgiskard-oss\u002Factions\u002Fworkflows\u002Fci.yml\u002Fbadge.svg?branch=main\" alt=\"CI\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgisk.ar\u002Fdiscord\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fdiscord\u002F939190303397666868?label=Discord\" alt=\"Giskard on Discord\" \u002F>\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>\u003Ca rel=\"nofollow ugc noopener\" href=\"https:\u002F\u002Ffosstodon.org\u002F@Giskard\">\u003C\u002Fa>\u003C\u002Fp>\n\u003C\u002Fdiv>\n\u003Ch3>\n   \u003Ca href=\"https:\u002F\u002Fdocs.giskard.ai\u002Foss\" rel=\"nofollow ugc noopener\">\u003Cb>Docs\u003C\u002Fb>\u003C\u002Fa> •\n  \u003Ca href=\"https:\u002F\u002Fwww.giskard.ai\u002F?utm_source=github&utm_medium=github&utm_campaign=github_readme&utm_id=readmeblog\" rel=\"nofollow ugc noopener\">\u003Cb>Website\u003C\u002Fb>\u003C\u002Fa> •\n  \u003Ca href=\"https:\u002F\u002Fgisk.ar\u002Fdiscord\" rel=\"nofollow ugc noopener\">\u003Cb>Community\u003C\u002Fb>\u003C\u002Fa>\n \u003C\u002Fh3>\n\u003Cbr \u002F>\u003Cblockquote>\n\u003Cp>[!IMPORTANT]\n\u003Cstrong>Giskard v3\u003C\u002Fstrong> is a fresh rewrite designed for dynamic, multi-turn testing of AI agents. This release drops heavy dependencies for better efficiency while introducing a more powerful AI vulnerability scanner and enhanced RAG evaluation capabilities. For now, the vulnerability scanner and RAG evaluation still rely on Giskard v2.\n\u003Cstrong>Giskard v2 remains available but is no longer actively maintained.\u003C\u002Fstrong>\nFollow progress → \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Forgs\u002FGiskard-AI\u002Fdiscussions\u002F2250\" rel=\"nofollow ugc noopener\">Read the v3 Announcement\u003C\u002Fa> · \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FGiskard-AI\u002Fgiskard-oss\u002Fissues\u002F2252\" rel=\"nofollow ugc noopener\">Roadmap\u003C\u002Fa>\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Ch2>Install\u003C\u002Fh2>\n\u003Cpre>\u003Ccode class=\"language-sh\">pip install giskard\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Cp>Requires Python 3.12+.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Telemetry:\u003C\u002Fstrong> Libraries built on \u003Ccode>giskard-core\u003C\u002Fcode> (including \u003Ccode>giskard-checks\u003C\u002Fcode>) may send \u003Cstrong>optional, aggregated usage analytics\u003C\u002Fstrong> to help improve the product. No prompts, model outputs, or scenario text are included. See \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FGiskard-AI\u002Fgiskard-oss\u002Fblob\u002FHEAD\u002Flibs\u002Fgiskard-core\u002FREADME.md#telemetry\" rel=\"nofollow ugc noopener\">what is collected and how to opt out\u003C\u002Fa>.\u003C\u002Fp>\n\u003Chr \u002F>\n\u003Cp>Giskard is an open-source Python library for \u003Cstrong>testing and evaluating agentic systems\u003C\u002Fstrong>. The v3 architecture is a modular set of focused packages — each carrying only the dependencies it needs — built from scratch to wrap anything: an LLM, a black-box agent, or a multi-step pipeline.\u003C\u002Fp>\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>Status\u003C\u002Fth>\n\u003Cth>Package\u003C\u002Fth>\n\u003Cth>Description\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003C\u002Fthead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>✅ Beta\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>giskard-checks\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>Testing & evaluation — scenario API, built-in checks, LLM-as-judge\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>✅ Beta\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>giskard-scan\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>Agent vulnerability scanner — red teaming, prompt injection, data leakage (successor of \u003Ca href=\"https:\u002F\u002Flegacy-docs.giskard.ai\u002Fen\u002Fstable\u002Fopen_source\u002Fscan\u002Findex.html\" rel=\"nofollow ugc noopener\">v2 Scan\u003C\u002Fa>)\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>📋 Planned\u003C\u002Ftd>\n\u003Ctd>\u003Ccode>giskard-rag\u003C\u002Fcode>\u003C\u002Ftd>\n\u003Ctd>RAG evaluation & synthetic data generation (successor of \u003Ca href=\"https:\u002F\u002Flegacy-docs.giskard.ai\u002Fen\u002Fstable\u002Fopen_source\u002Ftestset_generation\u002Findex.html\" rel=\"nofollow ugc noopener\">v2 RAGET\u003C\u002Fa>)\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftbody>\u003C\u002Ftable>\n\u003Ch2>Giskard Checks — create and apply evals for testing agents\u003C\u002Fh2>\n\u003Cpre>\u003Ccode class=\"language-sh\">pip install giskard-checks\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Cp>\u003Cstrong>\u003Ca href=\"https:\u002F\u002Fdocs.giskard.ai\u002Foss\u002Fchecks\" rel=\"nofollow ugc noopener\">Giskard Checks\u003C\u002Fa>\u003C\u002Fstrong> is a lightweight library for creating evaluations (evals) that test LLM-based systems — from simple assertions to LLM-as-judge assessments. Unlike traditional unit \u003C\u002Fp>\n",1784240407276]