Custom agents

A session-scoped memory layer for LLMs, built on the Model Context Protocol. LatentContext gives AI assistants the ability to explicitly store and retrieve notes, decisions, and context within a single conversation — keeping the AI focused on the current task without cross-session data contamination.

Build and manage intelligent hybrid AI agent systems that combine multiple models to enhance automation and decision-making.

Airline customer service agent on AWS Bedrock AgentCore with dual-memory: conversation history + Solutions Memory for successful cases. Users click "Like" to save resolutions for future reference. Includes Gateway policy engine, MCP tools, and Cognito auth. Handles refunds, delay compensation, and loyalty rebooking.

AI-driven design optimization of a 3D-printable anti-glugging vortex funnel. Automates the loop: parametric CadQuery geometry → OpenFOAM CFD (interFoam VOF multiphase via Docker) → multi-fidelity Bayesian optimization (scikit-optimize) — driven end-to-end by a Claude Code skill.

This is a golden path (opinionated shortcut) to explore AI from local homelab to PROD using OSS for technological independence and local LLM/SLM for privacy. We explore concepts like AI agents, RAG, LLM/SLM and MCP arch using OSS technologies like Linux, Opentofu, microk8s, DAPR, Ollama, TimescaleDB & pgai Vectorizer, Flask, Chainlit and Python.

Kuzushi — Agentic SAST scanner with AI triage

AI Researcher & Automation Architect | PhD in Multimodal Learning | Building Solarpunk solutions through Agentic AI | Open to collaborations on research, books & business automation projects 🌱🤖

Read PDFs with LLMs — @mention threads on passages, or generate interactive chapter summaries

🔬 Claude Code toolchain

Steve is an AI agent. Steve wants a body. This is his website. 🤖☘️

Unified MCP server for all League of Legends Riot API endpoints.

Convert natural language to terminal commands using AI. Simply AI helper in the Terminal

A ready-to-use reveal.js template for academics who want to code their lecture slides in HTML instead of PowerPoint, Keynote, or Google Slides. Fork it, customise it, and present from your browser within minutes.

A Model Context Protocol server for Microsoft SQL Server

A visual builder that generates structured prompts for AI to create UI components. Build complex component hierarchies, configure props, and export AI-ready prompts with real-time validation.

Anti-Gravity: The Agentic OS that turns AI (Antigravity, Cursor, Codex, Claude Code, Gemini CLI, OpenCode, VS Code) into disciplined senior engineers. A 150+ file architecture featuring persistent memory, execution workflows, and state-of-the-art UI/UX generation.

A ready-to-deploy Kubernetes-based setup for an all-purpose AI assistant, coding agent and complete homelab based on OpenClaw.

🤖 Engage in dynamic conversations with multiple AI personalities in a group chat, enhancing your interactive experience with real-time responses.

JARVIS-level AI assistant with 70+ features: Voice control, Computer Vision, AR, Gesture recognition, Multi-provider LLM (6 providers), Bug bounty automation, Code generation, Business intelligence. 2-4x more powerful than ChatGPT!

An AI-powered portfolio analysis assistant for Zerodha Kite users, built to learn MCP + LLM + RAG + Agentic AI fundamentals

Self-hosted AI character roleplay system. 100% local, zero data leaks. Your conversations stay on YOUR computer. Open-source Character.AI alternative using Ollama.

This project aims to transform sports education by leveraging cutting-edge technologies like Artificial Intelligence and Machine Learning. It provides an interactive platform for athletes, coaches, and enthusiasts through real-time video conferencing, intelligent chatbots, personalized score prediction, and curated news feeds.

CatBot - An example chatbot application based on Hubot. ConfigCat is a hosted feature flag service: https://configcat.com. Manage feature toggles across frontend, backend, mobile, desktop apps. Alternative to LaunchDarkly. Management app + feature flag SDKs.

Echo AI is a web-based interactive chatbot that simulates and enhances human-like conversations through a customizable interface and flexible logic.

WhatsApp Agent with Mastra AI framework integration. Features persistent memory (PostgreSQL + pgvector), circuit breaker resilience, retry logic, and MCP tool support. Includes comprehensive Jest ESM testing.

Voice-first ChatGPT companion that fuses multilingual speech recognition with lifelike AI synthesis directly in your browser.

A simple Frontend for AI Chatbot | next.js | This can answer your questions.💬

🧬 灵魂存档 — 通过日常对话静默构建你的数字人格克隆体 | A silent engine that builds digital personality clones through everyday conversations

CraveCrafters is an AI-powered food ordering web application that seamlessly integrates a chatbot to assist users with menu browsing, order placement, and customer service. The system features a Node.js backend, a FastAPI-based chatbot, and an interactive frontend built with HTML, CSS, and JavaScript. It utilizes MongoDB as its database.

Sentique is a full‑stack feedback analytics platform that ingests user reviews from App Store, Google Play, Trustpilot, Reddit and Twitter/X, processes them with a fine‑tuned BERT model into 16 categories and sentiment labels, and provides comprehensive analytics and insights.

Automate your instant messaging apps with ease! Run and manage custom plugins, AI powered chatbot, scheduled tasks and more.

AI SMS Auto-Responder for Android. Turn your Android device into an autonomous AI communication hub. A Python-based SMS auto-responder running natively on Termux, powered by LLMs (OpenRouter/Ollama) with a sleek Web & Terminal UI.

Experimental extension to chat with models available to download from Ollama

An interactive learning platform for college classmates featuring an AI chatbot, flip cards, and topic-based resources to make studying easier and more engaging.

This course on AI for software development explores the use of AI large language models (ChatGPT, Bard, etc) and their potential benefits and challenges. Hands-on activities show the ways in which AI can speed up software development tasks and free up time for more creative and strategic work, maximizing benefits/efficiency while limiting harm.

A lightweight Retrieval-Augmented Generation (RAG) system with a modern web interface, built in Go.

Python-Godot bridge for AI character control powered by OpenRouter API. VRM model management, animation playback, expression triggers, event handling, and bidirectional socket/godot-rl communication.

AI-WeChat-Mater 是一个开源项目，旨在将人工智能能力深度集成到微信生态系统中，提供自动化消息处理和智能交互功能。该项目支持多微信账号管理、智能AI对话、定时任务执行和全方位数据分析。

DingTalk group chat bridge for Claude Code. Send messages, receive @mentions, auto-execute prompts via Stream SDK.

Um código que implementa machine learning e IA para criar um chatbox, como chatGPT, que responde perguntas relacionadas ao banco de dados que foi conectado. Isso demonstra o poder da IA + ML para reduzir tempo de procura por respostas simples e assim o foco pode ser em partes mais importantes. Na versão 1.0 é esperado que reduza até 15% do tempo

Expert Systems Principles and Programming, Third Edition-Course Technology

AI-powered Dental Screening and Consultation System — A full-stack solution combining deep learning for dental disease detection (X-rays & images), a FastAPI backend with chatbot support, and Flutter/React frontends for mobile and web. Focused on improving accessibility of dental care with intelligent analysis and recommendations.

A simple chatbot using Node.js, Express, and the Gemini AI API.

Experimental Research and Development in Software and Mechanical Engineering for Artificial Intelligence, Robotics, and Design

This uses chat gpt api to make a chat bot in c# that can rolplay as anything/anyone and answer anything

English Learning Support Application

Welcome to SocialChatBot, your multi-platform, personality-driven chatbot for Instagram and Snapchat. SocialChatBot isn’t just another bot—she’s a fictional 22-year-old art history student in Berlin, with a sharp wit, a unique style, and a memory for your conversations. She’s designed to make every chat feel real.

A local emotional AI chatbot built with Python + LLaMA3 via Ollama. Powered by memories, feelings, and conversation — an AI companion that truly remembers.

Chat with YouTube videos using AI - paste a URL, get transcripts, ask questions

This repository presents a project focused on developing a high-precision legal expert LLM application called Contract Advisor RAG. The project's goal is to create a Retrieval Augmented Generation (RAG) system for Contract Q&A, enabling users to interact with contracts by asking questions and receiving accurate, context-rich responses.

Hands-on LLM evaluation learning repo — local models via Ollama, no paid APIs, no maths. Covers deterministic eval, LLM-as-a-Judge, hallucination testing, prompt injection, RAG evaluation, and agent trajectory scoring.

AI decision workflows that turn enterprise documents into traceable decisions, executive artifacts and operational handoffs.

面向小说创作场景的半自动化辅助系统

Your private document brain. PDFs in, RAG out. Self-hosted. Plug everywhere.

🧠 Transform documentation chaos into a structured memory system with Mnemos, your self-hosted, multi-context knowledge server for developers.

Navigation index for the 131 S-class problems forming the backbone of the WFGY tension reasoning engine.

Agent skill that teaches AI agents how to manage Zilliz Cloud vector databases through zilliz-cli

Simple local RAG sample using Spring-AI, Ollama LLMs and Vespa-AI

Turn your PDF collection into clean text your AI can actually read. Mac app.

微信公众号文章总结器 | AI-powered WeChat article summarizer with RAG, GraphRAG, MCP, and multi-platform export

Semantic search over your Gmail, from inside Claude Code. Embeds your full inbox with zembed-1 and reranks with zerank-2 — finds emails by meaning, not keywords.

Hybrid RAG system for translating Ancient Egyptian transliterations using the TLA dataset.

Wiii is an open-source durable AI workbench for conversations, local and cloud agents, files, tools, memory, artifacts, and permission-aware integrations.

An intelligent local AI agent powered by open-source LLMs, featuring free web search, hybrid memory, and context-aware query rewriting for real-time, grounded answers.

A lightweight wrapper around AI SDK that intelligently indexes and retrieves MCP tools using graph-based vector search.

Distributed document RAG system with intelligent GPU/CPU orchestration. Auto-discovers heterogeneous nodes, routes workloads adaptively, and achieves 60x+ speedups through VRAM-aware load balancing. Privacy-first architecture with 4 interfaces (CLI, API, MCP, Web UI). Real distributed systems engineering, not just an API wrapper.

Smart AI Based Food RAG is an intelligent food search system that utilizes cutting-edge technologies to provide users with efficient and intuitive text and image search capabilities. The system combines the power of Retrieval Augmented Generation (RAG) techniques, computer vision.

A high-performance, private Markdown search engine with Hybrid Search (BM25 + Vector) and MCP support. Built in Rust.

一个基于 Spring Boot + Vue 3 的现代博客系统，支持 AI 辅助写作、内容总结和 RAG 知识库系统

LLM powered Mental Health Support

A standalone Edge-RAG Memory Daemon for autonomous AI agents. High-performance Rust core (SurrealDB/RocksDB + Axum) with Bayesian Math Retrieval, multi-language AST indexing, and enterprise Semantic Redaction.

A Genshin Impact Question Answer Project supported by Qwen1.5-14B-Chat

The industry's first "Open Source OCR Arena," a free, no-login utility for one-click benchmarking of 7 top-tier models (Marker, MinerU, MonkeyOCR, Docling, Dolphin, OCRFlux, PP-StructureV3) on your PDF/image files, specializing in PDF-to-Markdown conversion.

[ICMR 2026] Official Repository for The Paper, Zero-Effort Image-to-Music Generation: An Interpretable RAG-based VLM Approach

Question-answering system for Zotero libraries with a 3-stage RAG pipeline: FAISS retrieval → CrossEncoder reranking → extractive QA. Features GROBID parsing, question expansion, multi-color PDF highlighting, and Streamlit interface. Ask questions, get answers with sources.

Helix: AI-powered CLI with RAG intelligence that converts natural language to commands using 450+ system commands. Features local AI, cross-platform package management, Git workflows, syntax highlighting, and multi-layer safety with directory sandboxing.

Turn your voice into intelligent, linked notes inside Obsidian

RAG-less document QA that loads an entire document into the LLM KV cache once, saves it to disk, and restores it before every query. No embeddings, no vector DB, no chunking. Drastic latency cut on static corpora.

RAG document intelligence with .NET AI building blocks (Microsoft.Extensions.AI, VectorData, Agents) and Aspire

Vex - Vector Exchange - Built by Vektor Memory - A Cross-standard Vector memory DB migration tool - Open Source

A multimodal RAG application using Qwen 2.5 VL, ColPali, and QdrantDB for text and image-based retrieval.

Fast hybrid (BM25 + semantic) local code search for AI agents - pure Rust, persistent index, MCP/gRPC servers, tree-sitter symbols

Partition API used by Agentset

Agentic AI Platform for intelligent database interaction

Local-first context engine for coding agents: project skeleton, symbols, references, graph, retrieval, confidence, MCP, CLI, and HTTP API.

一个为 Dify Workflow “量身定做”的 Telegram 群聊机器人。

Musubi (結び) — Ai Agent shared memory and thought layer. The braiding of threads between presences.

Cognitive Memory Architecture for Agents

A Columbia University capstone project focused on mitigating hallucinations in Medical Question Answering systems using Retrieval-Augmented Generation (RAG), ElasticSearch, and LLM-based validation.

A fully functional conversational AI system for restaurant phone ordering using Twilio, Whisper, LLM, and RAG

Full Stack Developer | AI & RAG Expert | Open Source Contributor. Bridging the gap between robust engineering and cutting-edge AI.

This MCP server enables LLMs to retrieve and process web scraping requests using ScraperAPI.

Official GitScrum MCP Server: Full Agile & CRM context for AI Agents in Cursor, VS Code and Claude.

AI-CLI 时代的 data lens:把 Claude Code 的 jsonl 实时镜像成可读 HTML,带跨 session 搜索 / 知识图谱 / 自动周报。零 token,单文件 Python。

Local-first MCP gateway for federating stdio and HTTP servers through one endpoint

A PowerShell tool for Windows that recursively scans a directory, captures its structure, and concatenates file contents. Perfect for creating detailed context for Large Language Models (LLMs) from your codebase.

Pi coding agent extension for native Ollama — fixes tool calling under streaming

This will be used as the repository for the demo with x402 and fastset

Curated OpenClaw/Codex skills collection for Apollo Toolkit.

Your Gemini web app as a service — turn the free Gemini web interface into a programmable HTTP API for image and video generation