[{"data":1,"prerenderedAt":4},["ShallowReactive",2],{"readme:all-agentic-architectures":3},"\u003Cdiv align=\"center\">\u003Cbr \u002F>\u003Ch1>Agentic Architectures\u003C\u002Fh1>\n\u003Ch3>Thirty-five production-grade agentic AI patterns. End to end.\u003C\u002Fh3>\n\u003Cp>A library \u003Cem>and\u003C\u002Fem> a living textbook — real LLM outputs, provider-agnostic,\ndeterministic-picker discipline throughout, and a comparative benchmark\nleaderboard that ranks every architecture against every relevant task.\u003C\u002Fp>\n\u003Cbr \u002F>\u003Cp>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FFareedKhan-dev\u002Fall-agentic-architectures\u002Factions\u002Fworkflows\u002Fci.yml\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002FFareedKhan-dev\u002Fall-agentic-architectures\u002Fci.yml?branch=main&label=CI&logo=githubactions&logoColor=white&style=for-the-badge&color=0a0a0a\" alt=\"CI\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Ffareedkhan-dev.github.io\u002Fall-agentic-architectures\u002F\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002FFareedKhan-dev\u002Fall-agentic-architectures\u002Fdocs.yml?branch=main&label=DOCS&logo=materialformkdocs&logoColor=white&style=for-the-badge&color=6366f1\" alt=\"Docs\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fpypi.org\u002Fproject\u002Fagentic-architectures\u002F\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fv\u002Fagentic-architectures?style=for-the-badge&logo=pypi&logoColor=white&label=PyPI&color=a855f7\" alt=\"PyPI\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FFareedKhan-dev\u002Fall-agentic-architectures\u002Fblob\u002FHEAD\u002FLICENSE\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Flicense-MIT-ec4899?style=for-the-badge&logo=opensourceinitiative&logoColor=white\" alt=\"License\" \u002F>\u003C\u002Fa>\u003C\u002Fp>\n\u003Cbr \u002F>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FFareedKhan-dev\u002Fall-agentic-architectures\u002Fblob\u002FHEAD\u002Fdocs\u002Fgetting-started\u002Fquickstart.md\" rel=\"nofollow ugc noopener\">\n  \u003Cimg alt=\"Quickstart\" src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FQuickstart-→-0a0a0a?style=for-the-badge&labelColor=0a0a0a\" \u002F>\n\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FFareedKhan-dev\u002Fall-agentic-architectures\u002Fblob\u002FHEAD\u002Fdocs\u002Findex.md\" rel=\"nofollow ugc noopener\">\n  \u003Cimg alt=\"Documentation\" src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDocumentation-→-262626?style=for-the-badge&labelColor=262626\" \u002F>\n\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FFareedKhan-dev\u002Fall-agentic-architectures\u002Fblob\u002FHEAD\u002Fdocs\u002Farchitectures\u002Findex.md\" rel=\"nofollow ugc noopener\">\n  \u003Cimg alt=\"Architectures\" src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FArchitectures-→-404040?style=for-the-badge&labelColor=404040\" \u002F>\n\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FFareedKhan-dev\u002Fall-agentic-architectures\u002Fblob\u002FHEAD\u002Fdocs\u002Fbenchmarks.md\" rel=\"nofollow ugc noopener\">\n  \u003Cimg alt=\"Benchmarks\" src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FBenchmarks-→-525252?style=for-the-badge&labelColor=525252\" \u002F>\n\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fcodespaces.new\u002FFareedKhan-dev\u002Fall-agentic-architectures\" rel=\"nofollow ugc noopener\">\n  \u003Cimg alt=\"Open in Codespaces\" src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FOpen_in_Codespaces-→-737373?style=for-the-badge&labelColor=737373\" \u002F>\n\u003C\u002Fa>\u003Cp>\u003Cbr \u002F>\u003Cbr \u002F>\u003C\u002Fp>\n\u003Ctable>\n\u003Ctr>\n\u003Ctd align=\"center\">\u003Ch2>\u003Ckbd>  35  \u003C\u002Fkbd>\u003C\u002Fh2>\u003Csub>ARCHITECTURES\u003C\u002Fsub>\u003C\u002Ftd>\n\u003Ctd align=\"center\">\u003Ch2>\u003Ckbd>  283  \u003C\u002Fkbd>\u003C\u002Fh2>\u003Csub>PASSING TESTS\u003C\u002Fsub>\u003C\u002Ftd>\n\u003Ctd align=\"center\">\u003Ch2>\u003Ckbd>  17  \u003C\u002Fkbd>\u003C\u002Fh2>\u003Csub>BENCHMARK TASKS\u003C\u002Fsub>\u003C\u002Ftd>\n\u003Ctd align=\"center\">\u003Ch2>\u003Ckbd>  9  \u003C\u002Fkbd>\u003C\u002Fh2>\u003Csub>LLM PROVIDERS\u003C\u002Fsub>\u003C\u002Ftd>\n\u003Ctd align=\"center\">\u003Ch2>\u003Ckbd>  0  \u003C\u002Fkbd>\u003C\u002Fh2>\u003Csub>MOCKED RUNS\u003C\u002Fsub>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\u003Cbr \u002F>\u003C\u002Fdiv>\u003Chr \u002F>\n\u003Ch2>Overview\u003C\u002Fh2>\n\u003Cp>A single Python library that packages every major agentic AI pattern from the literature as a runnable \u003Ccode>Architecture\u003C\u002Fcode> class with a uniform contract. Each pattern ships with a fully executed Jupyter notebook whose theory is written \u003Cem>against\u003C\u002Fem> the captured run — not synthetic examples. The library is multi-provider (Nebius, OpenAI, Anthropic, Groq, Ollama, Together, Fireworks, Mistral, Google) and built on top of LangGraph state machines.\u003C\u002Fp>\n\u003Cp>The central technical discipline of the repository is the \u003Cstrong>deterministic-picker pattern\u003C\u002Fstrong> — every LLM-as-Scorer surface has the LLM commit to categorical features (booleans, enums) and lets Python compose the deciding signal. This is the universal escape from the LLM-as-Scorer flat-band pathology, applied in 13 of 35 architectures; 9 more are architecturally immune by design.\u003C\u002Fp>\n\u003Chr \u002F>\n\u003Ch2>Quickstart\u003C\u002Fh2>\n\u003Cpre>\u003Ccode class=\"language-bash\">pip install \"agentic-architectures[nebius,faiss,tavily]\"\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Cpre>\u003Ccode class=\"language-python\">from agentic_architectures import get_llm\nfrom agentic_architectures.architectures import Reflection\n\narch = Reflection(llm=get_llm(), max_iterations=2, target_score=8)\nresult = arch.run(\"Write a haiku about a glacier.\")\n\nprint(result.output)\nprint(\"score:\", result.metadata[\"final_score\"], \"\u002F 10\")\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Cp>Same \u003Ccode>.run(task)\u003C\u002Fcode> interface across all 35 architectures. Same \u003Ccode>ArchitectureResult\u003C\u002Fcode> return shape. Swap the class, swap the patter\u003C\u002Fp>\n",1784240406337]