Custom agents

Offline-first project context and durable team management for coding agents. A local SQLite CLI for tasks, sessions, checkpoints, worktrees, and read-only team context.

🔨Reusable personal Agent Skills and practice-backed Q&A for AI Agent engineering, runtime reliability, multi-agent systems, evaluation, and AI-native frontend.

Toolkit for AI whitehats, internal red teams, llm bug bounty hunters and mlops - Adversarial testing for AI, LLMs, and Agents.

🤖 Advanced multi-agent robotics system using ROS2, NVIDIA Isaac Sim, behavior trees, and RL locomotion controllers. Features mobile manipulator, forklift, and Spot robot with autonomous navigation and manipulation. Graduation Project 2025.

对Godot开发游戏适配的Agent客户端，基于Pi开发。The Agent client for adapting games developed with Godot, based on Pi

A platform for running controlled experiments on teams of LLM agents: multi-agent simulations, an event-sourced run log, swap and fork interventions, and post-hoc evaluation metrics.

基于 EdgeOne 的自托管免费微信推送与标准 MCP Server，支持 Agent / IDE、Webhook 风格 API、多应用和订阅群发。Self-hosted free WeChat push service and standard MCP server on EdgeOne, with Agent/IDE support, webhook-style API, multi-app support, and subscription broadcast.

Full-SDLC suite of 57 agent skills for Claude Code & Codex: business analysis → system architecture → TDD development → review → QA → release. Neo4j knowledge graph as the source of truth for every requirement.

Purpose-built AL Copilot Skills that help agents implement consistent Business Central patterns and development best practices.

Control an Android device's shell from AI clients through MCP - outbound-only, no root or Shizuku.

asAgent brings chat, browser assistance, scheduled tasks, and reversible file operations into one desktop app.

A zero-dependency Python tool to resolve relative Markdown links and inline images as Base64 strings for RAG and Multimodal LLMs.

Guardrails and observability for autonomous AI agents - tool-call auditing, spend caps, and rollback for anything an agent does in production.

Guardrails and observability for autonomous AI agents - tool-call auditing, spend caps, and rollback for anything an agent does in production.

Cogito, ergo sum. A portable self-awareness framework for LLM agents — TICK heartbeat, Focus Stack, Temporal parsing, Self-Perception. Platform-agnostic.

Code for our paper: "Exploring Design of Multi-Agent LLM Dialogues for Research Ideation", SIGDIAL2025

Writ: An AI agent skill that catches AI writing tells and clichés using a self-edit checklist drawn from Wikipedia's Signs of AI Writing guide.

Skills for running, porting, profiling, and tuning code on Intel CPUs, GPUs, and the libraries and runtimes behind them.

Course Compass 学生选课智能体 | RAG · Function Calling · 约束排课 · Streamlit 本地 Demo

The agent that evolves with you 𖤍

Open platform for running Managed Agents at scale

Moku — The pocket cockpit for the vibe-coding era. Auto-resumes 12+ AI coding CLIs (Claude Code, Codex, OpenCode), Rust terminal engine (28 MB/s), Mosh roaming, Dynamic Island pixel pet.

Lightweight, zero-bloat multi-agent AI engine in a single daemon. Homelab-friendly and built for mini devices (Raspberry Pi, Orange Pi, Geekom, NUC, LXC).

[ICKG] Survey resources for personalized graph memory for llm agents.

A research terminal for prediction markets, with multi-agent evidence gathering, probability maps, and inspectable answers.

An intelligent AI agent system powered by LLM, designed to be a proactive personal assistant.

The desktop AI agent for the real web. Built-in fingerprint browsers, saved logins, multi-agent orchestration, and reusable Browser Skills beyond local coding.

A collection of practical Claude Code skills for real-world development workflows

STEER — AI4R (AI for Research) experiment framework for auto-research and autonomous research. Connect existing research projects and let an agent edit code and train under governance: frozen task, scorer, and evidence.

See your AI agents work. Real-time 3D visualization of Claude Code activity, multi-agent orchestration, and REST API integration for external systems.

A Java-implemented agent development framework

Open-source AI research assistant for VS Code + GitHub Copilot. Connects to PubMed, OpenAlex, Semantic Scholar, Europe PMC, CrossRef, and Zotero via MCP servers. Custom agents guide systematic reviews, academic writing, data analysis, and project management — all ICMJE-compliant with full audit trails.

AI가 쓴 한글 글의 AI 티를 제거하는 Kiro CLI 에이전트 (im-not-ai kiro ver.)

A curated list of Claude Code, Cursor, and Copilot skills and plugins built for students: IB, IGCSE, college-bound, and anyone studying with an AI coding agent at their side.

AI-powered financial analysis agent: real-time market data, multi-analyst workflows, and portfolio tools in your terminal.

Convention-over-configuration framework for multi-agent Claude Code development

OpenCodeReview (Community Edition) - Agentic AI based source code review tool powered by NVIDIA NeMo Agent Toolkit. NVIDIA Hackathon Award Winning Project

A local-first Windows AI assistant with a multi-model council pipeline, Python sandbox, web search, and cloud AI support. Free. Private. No subscriptions.

Gym-style environment for Warcraft III Legacy: stepping, observations and commands through an injected DLL

[Archived] AAOSP, predecessor to JibarOS. Early experiment in OS-level on-device AI (MCP + local LLM on Android 15). See JibarOS for the active project.

Started as a €70 used mini PC running an always-on AI agent (Hermes, n8n, Pi-hole, monitoring, Tailscale, encrypted backups). Now with a second stack next to it: a homelabbed GPU server for local AI (Qwen 27B and Flash-Next, image generation, Ollama, one API). All as code.

Self-hosted command center for AI coding agents (Claude Code, Codex CLI, GitHub Copilot CLI) — live dashboard, orchestration, and an MCP server for programmatic fleet control.

Run always-on messaging agents on Agent Substrate. A multi-tenant gateway holds the channel connections while each conversation runs as a suspendable actor, so presence stays warm and cognition suspends to reclaim compute. Reference implementation built on OpenClaw.

Skills that let Claude Code, GitHub Copilot, Cursor, Kiro and other coding agents work with OutSystems 11 apps via the Service Studio MCP server.

Make generic UI look consistent. No hardcoding, no ignoring your design system. Spec for AI-native design systems: tokens, slots and rules agents can discover and validate via MCP (coming in v2).

Autonomous software engineer. Researches, plans, builds, commits, fixes bugs, loops until done.

让想法，有下一步。Fluent 风格桌面待办、AI 规划与沉浸专注工具，基于 PySide6 / QML。

An intelligent AI agent system powered by LLM, designed to be a proactive personal assistant.

Litespeed — a local coding agent with multi-model workflows for your terminal and browser

A full-stack personal finance manager powered by Agentic AI, MCP, Next.js, Express, MongoDB, and a TypeScript Turborepo. Track accounts, transactions, budgets, goals, and analytics with a responsive UI, CSV import, and production-ready DevOps infrastructure (Docker, Kubernetes, Terraform, & multi-cloud support - AWS, Azure, OCI, and GCP)

Access Claude plugin marketplaces from Pi Coding Agent.

Claude Code plugin marketplace: repo-agnostic skills, hooks, agents, and MCP servers.

Modern MCP server for accessing Monarch Money financial data using AI agents, like Claude. Token-efficient, smart tool design, and uses latest MCP protocol features

The most complete gallery of what people build with Jev, TypeSafe's System One model: 3,400+ projects, demos, and write-ups by scenario, each with its original link, image, and description.

Autonomous coding agent orchestrator — polls GitHub Issues, runs agents in isolated worktrees, opens PRs

Local-first code intelligence for AI agents — tree-sitter symbol graphs and hybrid (FTS5 + vector + graph) retrieval via CLI and MCP server.

K.G.Studio is a lightweight, browser‑based DAW with an LLM‑powered AI Agent "K.G.Studio Musician Assistant" for AI‑assisted composition and MIDI editing.

AI 八字 + 紫微斗数排盘与综合印证 Skill：算法精准排盘（不靠 LLM 猜），三种分析模式，一键生成水墨风 HTML 命盘海报。兼容 Claude / Codex / Cursor / Workbuddy 等 SKILL.md Agent。

k8s_operation 是一个对标 Rancher/KubeSphere 的企业级 Kubernetes 后台管理系统，基于 Go + Gin + GORM + Vue3 + client-go + AI 构建。集成多集群管理、可视化资源操作、三层 RBAC 权限、CI/CD 发布编排、AI 智能助手（自然语言运维 / Function Calling / 多模型提供商 / 高危操作审批）

GPT-image-2 的 prompt 工程方法论。核心：用真实参考名替代形容词，prompt 默认中文短句。含单图 playbook、批量网格生成、失败模式树。跨 agent 通用。

Dify agent demo using Doubao and a Go/OpenAPI movie-search tool backed by Meilisearch.

本地优先 AI 设计 Agent 与无限画布 · 当前源码禁止未经授权商用 | Local-first AI design workspace. Source available; non-commercial use only without written permission.

Open-source deep research harness for evidence-grounded reports: multi-agent planning, independent section research, and targeted report revision.

Free, open-source Windows desktop toolbox for local file search, document previews, OCR batch renaming, and an optional AI assistant.

🤖 Каталог ИИ-инструментов для разработчиков на русском (2026): AI-IDE, Cursor, Copilot, Claude Code, Codex, агенты, MCP, код-ревью, тесты, LLMOps, локальные LLM, российские сервисы

This project is for learning: it shows how an agent loop works underneath, in a way more intuitive and visual than reading about it. Behind an agent's chat box runs a loop: step after step, the model proposes an action and the harness carries it out, until it stops.

CI for LLM agents — change the prompt, swap the model, edit a tool, refactor the loop: replay the eval suite and let statistics decide if anything got worse. OTel tracing, calibrated LLM-as-judge,adversarial simulated users, and a GitHub Actions gate that blocks significant regressions.

Code for the multi-agent computer use project.

面向售前工程师、解决方案架构师与技术销售的实战型知识库：方法论、政府/B端项目、AI大模型/RAG/Agent、架构设计、招投标、沟通汇报、模板框架、行业方案与竞品分析。

cost-governed multi-agent RAG over SEC 10-K filings

Turn a WeChat group's day into a verifiable illustrated report (PNG/HTML/Markdown) — evidence-bound editorial, read-only, offline-runnable examples

Agents for any host, composed with server primitives through one Vite plugin.

[EMNLP 2026 main] VideoSearcher: Empowering Video Deep Research with Multi-Tool Agentic Reasoning via Reinforcement Learning

Persistent Git-backed messaging for AI coding agents — coordinate Claude Code, Codex, and others across sessions, worktrees, and machines.

View planning with multi-turn VLM agents: ViewSuite 6-DoF benchmark on real ScanNet scenes + iterative RL-SFT training

Collection of frameworks and tools for AI evalations, including tool-use, agentic AI, MCP, and multimodal

A robot-oriented full-stack framework for long-horizon embodied agents in evolving physical and social environments.

非技术产品经理/学生从想法到上线的 AI 产品开发工具包（规则卡 + 三份技术手册 + 操作手册）

Multi-stage LLM agent pipeline for optimizing AI kernels on Intel GPU — from analysis to autotuning.

Implementation of the Decrypto benchmark for multi-agent reasoning and theory of mind.

Agent — A next-generation sandbox for AI code agents. Environment management, CI/CD pipelines, and multi-agent collaboration.

面向个人用户的轻量化健康管理工具，基于LLM的 Agent 技术，解决用户日常用药安全查询、化验单快速解读、个人就诊记录统一归档的核心需求，全程严格遵守医疗合规要求，不提供任何诊疗服务。

Prism is a windows desktop ai agent built with electron and a fastapi backend, powered by the google-gemini-3-api. it can understand what is on your screen, plan multi step actions.

let LLMs control your desktop for tasks

Harnessing AI to Disrupt and Evaluate Security (HADES)

A seq2seq based chatbot built with PyTorch (中文聊天机器人)

Template for building a chatbot using TypeScript and Next.js, powered by the Hugging Face API

SOC Analyst Level 1 Replacement using RAG LLM

Spinup your own LLM powered chat agents for solana actions

Multi-Agent LLM System that automates news gathering and blog post writing.

Fully Cooperative Multi-Agent Deep Reinforcement Learning

https://github.com/CryptoMonkey21/AgentGPT.git

A collection of 19 generative AI projects in Python, showcasing applications in text generation, image synthesis, and chatbots using frameworks like Transformers and PyTorch. Includes datasets, code, and tutorials for building and deploying cutting-edge AI models.

Competition and Agent Frameworks for the Trading Agents Competition

Sample dockerized NestJs framework service integration with New Relic Node.js Agent showing how to instrument the app in order to use all the Agent’s features (like Service Maps or Distributed Tracing)

🎮 Tetris with AI vs AI mode, 8-bit sound, theme shop, replay sharing & more — zero dependencies PWA

A powerful, modular, and self-hosted AI assistant that combines website monitoring, ERP integration, system health checks, web search, and personal organization into a single Telegram bot and Web Interface. Powered by Ollama for local, private intelligence.

An abstraction library for building domain-specific intelligent agents based on Large Language Models (LLMs). LLMAgent provides a core architecture and behavior definitions that simplify the development of specialized agents by leveraging the AgentForge framework for dynamic workflow orchestration.