[NeurIPS 2023] Reflexion: Language Agents with Verbal Reinforcement Learning

Reflexion: an autonomous agent with dynamic memory and self-reflection

Implemented various AI algorithms in Pac-Man projects developed by UC Berkeley. Implemented informed/blind state-space search using search algorithms like BFS, DFS, UCS and A* algorithm with heuristic calculation. Designed an algorithm for reflex agent, minimax and alpha-beta pruning. Reinforcement Learning using MDP (Value Iteration and Policy Evaluation) - Implemented Markov Decision process using Value iteration in the Pacman World. Designed perceptron classifier for data classification. Developed applications using python. http://ai.berkeley.edu/project_overview.html

A LangGraph template for building AI agents with human-in-the-loop, conditional routing, observability (Langfuse), and a Reflex UI demonstrated through a travel planner use case.

An implementation of a reflexion agent

A simple, modular, non-brittle framework for implementing text-based Reflexion agents.

An open-source security proxy and active firewall for the Model Context Protocol (MCP). It acts as a real-time 'Reflex System' between AI agents and their tools, protecting the host system from malicious intent, prompt injection, and data exfiltration.

An AI agent for a 5-card poker game. The agent is designed as a reflex agent, meaning it acts based on the strength of its hand without considering the opponent's bidding. The underlying strategy is to use machine learning, game theory, and decision-making algorithms to adapt and improve the agent's performance over time.

Vacuum Cleaner - Simple Reflex Agent

A simple reflex agent is a type of intelligent agent that performs actions based solely on the current situation. I desinged a reflexx agent to solve vaccuum cleaner problem.

Reflexion Agent implementation based on the smolagents library

Comprehend code

A modular Python implementation of a Simple Reflex Agent, designed to simulate intelligent behavior using condition-action rules. This project serves as an educational demonstration of one of the fundamental agent architectures in artificial intelligence.

A proof of concept for Continuous Learning in AI Agents using LangGraph and Reflexion technique. The agent learns from mistakes and transfers knowledge to new tasks via long-term memory. 🧠🤖

Reflex Agent Improvement, Minimax Search, Alpha-Beta Pruning, Expectimax Search, Custom Evaluation Function for Pacman.

Self-correction engine for AI coding agents - implements the Reflexion paper (Shinn et al., 2023) as a Rust MCP server. Turns failures into persistent, searchable lessons that prevent the same mistakes across sessions.

Simple reflex agent (Tree+Implement)

Reflex agent that solves the vacuum cleaner problem.

Open-source platform for building, deploying, and improving AI agent skills. Correction-driven feedback loop — every user correction becomes a test, a guardrail, and a reflexion rule that improves future executions.

Assignment 2: Reflex Agent, Minimax, Alpha-Beta Pruning, Introduction to Artificial Intelligence, Spring 2017, National Chiao Tung University

Symbolic reflex governance layer for LLM agents. Includes ACH-P and TRUSTFORM-1 protocols with ritual hooks for Archon/MemoryOS stacks.

A simulation for an automated vacuum cleaner in Unity. This project implements a simple-reflex agent to create a simulation of a vacuum that will go clean up the environment.

The reflex agent for polling inputs and running playbook actions.

Reflexion Agent using LangGraph, Tavily Search engine, and Function Calling

Artificial Intelligence Gomoku game, allows human beings playing against the agent. Contains the simplest Reflex Agent, the MiniMax Agent with and without alpha-beta pruning, the Stochastic Agent and the Supervised Agent.

A professional Python-based laboratory for designing, simulating, and evaluating rational AI agents, including simple reflex, model-based, utility-based agent architectures etc.

An autonomous research agent that synthesizes detailed reports by combining dynamic planning, unsupervised topic discovery (PCA/KMeans) HyDe (hypothetical doc embeddings), and a self-correcting reflexion loop.

Reflexion agents implemented on LangGraph.

An intelligent, RL-powered coding copilot with autonomous reasoning and a sleek dashboard.

Python AI Agents Hub: Your gateway to intelligent agent development in Python!