# FareedKhan-dev/all-agentic-architectures

35 production-grade agentic AI architectures (Reflexion, LATS, GraphRAG, MemGPT, Voyager, BrowserAgent, ...) — a Python library and runnable textbook with multi-provider LLM support and a 17-task benchmark leaderboard.

Repository: https://github.com/FareedKhan-dev/all-agentic-architectures
Canonical: https://ross.abutalabs.com/products/all-agentic-architectures
Homepage: https://fareedkhan-dev.github.io/all-agentic-architectures/
Language: Jupyter Notebook
License: MIT
License Family: permissive
Topics: agentic-ai, ai-agents, langchain, langgraph, langsmith, llm
Last push: 2026-06-22T06:16:53+00:00

## Health v2 (maintenance only)
Score: 63/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 88, release rhythm 54, longevity 24
- inputs: {"age_days": 343, "days_push": 72, "days_rel": 97, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4154, forks 732 (observed 2026-08-28T04:08:36.505561+00:00)

## What it is
A Python library packaging 35 production-grade agentic AI architectures (Reflexion, LATS, GraphRAG, MemGPT, Voyager, etc.) as runnable Architecture classes with a uniform contract, built on LangGraph. It doubles as a living textbook with executed Jupyter notebooks and includes a 17-task benchmark leaderboard comparing patterns across 9 LLM providers.

## Use cases
- implement agentic AI patterns like Reflexion or Tree of Thoughts in Python
- compare agent architectures on a benchmark before choosing one
- learn how GraphRAG or MemGPT work with real executed examples
- build a multi-provider LLM agent without vendor lock-in
- study agentic RAG variants like Corrective RAG and Self-RAG
- run a benchmark leaderboard of agent patterns across tasks
- add memory or self-critique loops to an LLM application

## When to choose
- you want runnable, tested implementations of many agentic patterns in one library
- you need provider-agnostic LLM support across OpenAI, Anthropic, Groq, Ollama, and others
- you are learning agent architectures and want theory backed by real captured runs
- you want to benchmark which agent pattern suits your task family

## When to avoid
- you need a production agent orchestration platform rather than reference implementations
- you want a no-code or GUI agent builder
- you only need a single simple agent without studying architecture trade-offs
- you cannot provide API keys or run LLM calls, since examples execute real models

## Facets
- artifact type: library
- maturity: active
- function: agent-framework, rag, llm-inference, benchmarking, machine-learning, prompt-engineering, chatbot
- domain: artificial-intelligence, large-language-models, machine-learning, developer-tools, tutorials
- platform: python, cross-platform, cli
- tags: agentic-ai, langgraph, langchain, jupyter-notebook, llm-providers, reflection-patterns, graphrag, memgpt, tree-of-thoughts, benchmark-leaderboard, textbook, multi-provider-llm, ai-agents, retrieval-augmented-generation

## Member repositories
- FareedKhan-dev/all-agentic-architectures (main) score 63

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:36.505561+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:22:52.548204+00:00, confidence not recorded.
  - readme: https://github.com/FareedKhan-dev/all-agentic-architectures (fetched 2026-08-28T04:08:36.505561+00:00, sha 6b13e82bfd1f)
  - homepage: https://fareedkhan-dev.github.io/all-agentic-architectures/ (fetched 2026-08-29T09:13:41.827114+00:00, sha 69e9e33ff249)
- Data as of 2026-08-30T08:39:29.467469+00:00.