Generative Agents

Top Builders

Explore the top contributors showcasing the highest number of app submissions within our community.

Generative Agents

Generative Agents are computer programs designed to replicate human actions and responses within interactive software. To create believable individual and group behavior, they utilize memory, reflection, and planning in combination. These agents have the ability to recall past experiences, make inferences about themselves and others, and devise strategies based on their surroundings. They have a wide range of applications, including creating immersive environments, rehearsing interpersonal communication, and prototyping. In a simulated world resembling The Sims, automated agents can interact, build relationships, and collaborate on group tasks while users watch and intervene as necessary.

General
Relese date	April 7, 2023
Type	Autonomous Agent Simulation

Start building with Generative Agents

We have collected the best Generative Agents libraries and resources to help you get started to build with Generative Agents today. To see what others are building with Generative Agents, check out the community built Generative Agents Use Cases and Applications.

Generative Agents Tutorials

Building an Intelligent AI Agent for Content Moderation with Structured Output

👉 Discover more Generative Agents Tutorials on lablab.ai

Generative Agents resources

Kickstart your development with Generative Agents.

Edit on GitHub

Generative Agents AI technology page Hackathon projects

Discover innovative solutions crafted with Generative Agents AI technology page, developed by our community members during our engaging hackathons.

SystemQuest

SystemQuest is a hands-on, gamified learning platform that transforms how engineers learn system design. Instead of passively reading blogs or memorizing patterns, users actively build real-world system architectures on a visual drag-and-drop canvas using components like load balancers, caches, databases, queues, and services. Each design is evaluated in real time by a simulation engine across core system metrics such as latency, throughput, availability, scalability, and cost. Users receive transparent, rubric-based feedback that highlights architectural gaps and trade-offs. An integrated AI assistant understands both the current mission and the user’s architecture, providing contextual explanations, hints, and design guidance at multiple levels of depth. SystemQuest offers structured learning paths and missions inspired by real engineering problems—ranging from high-read systems and asynchronous processing to distributed consistency and large-scale streaming. By combining simulation, AI feedback, and gamification, SystemQuest helps learners build true system design intuition through practice and iteration.

PulseForge

PulseForge is a production-ready multi-agent AI trading intelligence system built on Complete.dev. Instead of relying on a single black-box model, PulseForge operates as a coordinated network of specialized agents responsible for data ingestion, structured memory indexing, signal generation, deterministic trade simulation, and reward-based validation. The system continuously fetches market data, news, and public signals, stores them in a Supabase-backed memory layer, and generates regime-aware trading signals with structured explanations. Each simulated trade is executed using deterministic risk logic (stop-loss, take-profit, slippage, and fees), then scored by a validator agent using accuracy, risk efficiency, drawdown control, and confidence calibration. These validation rewards are fed back into adaptive policy weights, allowing the system to evolve strategy selection across different market regimes (trend, range, volatility). A single Manager endpoint orchestrates the full intelligence loop: Fetch → Store → Signal → Execute → Validate. PulseForge demonstrates how collaborative AI agents can form a self-evaluating, adaptive decision engine — transforming trading signals into structured, improving intelligence.

ClawScholar- Autonomous Research Capital Engine

ClawScholar is an autonomous research capital engine built using OpenClaw to demonstrate multi-step reasoning, persistent memory, and tool-driven execution in scientific workflows. Modern research ecosystems suffer from fragmented literature, inefficient capital allocation, and opaque funding decisions. ClawScholar addresses this by autonomously synthesizing academic papers, identifying theoretical contradictions and research gaps, and generating structured analytical insights. The system operates in tiered intelligence modes (Basic → Enhanced → Advanced → Elite), dynamically scaling analytical depth. It integrates a simulated on-chain treasury mechanism on the Sepolia testnet to demonstrate programmable research funding logic based on analytical confidence and complexity. Built with Streamlit, OpenClaw agent orchestration, and blockchain interaction layers, ClawScholar showcases how autonomous agents can evolve from passive assistants into active research decision engines. This project represents a prototype foundation for autonomous scientific infrastructure, where AI agents can evaluate knowledge, allocate resources, and support transparent, programmable research ecosystems.

VelocityHire

VelocityHire is an AI-powered recruitment automation platform designed to eliminate the inefficiencies of high-volume hiring. At its core, it leverages a 3-agent LangGraph pipeline that automates the entire recruitment workflow from candidate discovery to outreach. Agent 1 (Profile Analyzer) ingests candidate data and scores profiles across a 7-node graph, weighing factors like hackathon performance, technical skills, certifications, and recency to produce a ranked candidate score. Agent 2 (Job Matcher) takes scored profiles and runs them through a 5-node matching graph, evaluating adaptability, role fit, and culture fit to generate precise match scores against open positions. Agent 3 (Outreach Coordinator) closes the loop by segmenting matched candidates into PRIORITY, STANDARD, NURTURE, or ARCHIVE tiers and automatically generating personalized outreach campaigns tailored to each segment. Built on FastAPI, SQLAlchemy ORM, and a multi-tenant SQLite database, VelocityHire supports full organizational isolation via X-Company-ID scoping across 5 core tables. It integrates natively with leading ATS platforms including Greenhouse, Lever, and BambooHR via webhook endpoints, enabling seamless adoption into existing hiring stacks. The result is a scalable, intelligent hiring engine that dramatically reduces time-to-hire while improving candidate quality and engagement.

Ladybug - The Robot Reader

There are 240 million children worldwide living with learning disabilities, and many struggle to access physical books independently. Ladybug: The Robot Reader was built to change that. Ladybug is an autonomous robotic system that reads physical books aloud from cover to cover with no human intervention. Built on the SO-101 robotic arm, it uses a perception-action loop powered by Claude Vision to assess the workspace, decide what to do next, and execute — opening a closed book, reading each page spread, turning pages, and closing the book when finished. Claude Vision analyzes camera frames to classify page types (content, title page, table of contents, index, blank) and extract text. ElevenLabs then streams natural-sounding speech in real time using a sentence-level prefetch pipeline so audio plays continuously without pauses. Motor skills — opening, closing, and page turning — are trained using ACT (Action Chunking with Transformers) policies. The system includes intelligent retry logic with frame hashing to detect failed page turns and automatically retry them. Ladybug supports multiple reading modes: verbose (reads everything), skim (headers and titles only), and silent (text extraction only). It also features a web dashboard for remote monitoring and a dry-run mode for testing without hardware. Our mission is accessibility in education — putting an autonomous reading companion in every special education classroom. We want 1,000,000 lady bug robot readers available to children around the world.

ClutterBot

ClutterBot is a proof-of-concept simulation platform that bridges natural language understanding and robotic task execution for household cleanup tasks. Users issue commands like "pick up the phone and the toy train," which Gemini 3 Flash parses into structured task lists. The system generates complete execution plans upfront, with Gemini deciding the sequence of pick-and-place operations for each object. The architecture combines a FastAPI backend hosted on Vultr (central system of record), a Next.js frontend for real-time monitoring, and a MuJoCo physics simulation featuring a Franka FR3 manipulator in a room environment with everyday objects. The robot executes inverse kinematics motions to relocate objects from scattered positions on a table to a collection bin, with each action streamed via WebSocket for live visualization. This prototype validates the feasibility of integrating large language models with robotic simulation pipelines, demonstrating how AI can translate high-level human intent into executable robot behaviors. While the current implementation uses deterministic motion planning with hardcoded inverse kinematics rather than learned policies, the framework establishes foundational patterns for future work incorporating adaptive control, real hardware integration, and expanded object manipulation capabilities. The plan-first approach (Gemini generates the full task plan in a single API call) shows AI reasoning while keeping execution fast and deterministic, making it suitable for real-time interactive use.