Custom agents

Durable, private memory for AI agents. Fully offline MCP memory layer: LMDB + Tantivy/BM25, no models, no cloud. Ships with full source - audit it, extend it, make it yours.

Spring Boot + MyBatis 框架感知代码地图：静态分析调用链/SQL/事务/实体影响面，通过 MCP 喂给 AI 编程工具（Trae/Cursor/Claude Code），省 token、防幻觉

Autonomous personal AI agent.

极简 AI Agent 框架，支持多租户，适用于企业数字员工。

Adyber AI is an open-source AI desktop assistant that understands natural language and helps you perform everyday tasks. Open apps, search the web, answer questions, play music, and automate actions on your computer, all from one assistant.

面向公务员、事业单位等招考场景的 AI 数据分析工作台，支持招考数据导入、清洗、数据中心查看与智能岗位分析。

无人值守的端到端自动履约链路：发卡内核 + 不可再生库存号池 + 判据驱动的浏览器自动化

AI-powered post-exploitation framework — GitHub Gist dead-drop transport, no open port required.

Block cookie consent dialogs in Playwright with a single line of code. Uses 15,000+ rules from I Still Don't Care About Cookies.

Yad: a free, open-source, local-first browser agent. Chrome extension (Part A, the hand) + local Companion (Part B, the brain). Your API key is encrypted on your device, capped, and stoppable with one click.

AutoPi: desktop and command-line AI agent with event-driven Wake automation, derived from Pi.

CDP-native MCP server that lets an AI agent drive Chrome at full depth — deep network/console read, robust interaction, multi-session, and API-flow → runnable-code extraction.

AI Native Security Tool for Wallet && Password management.

Provider-agnostic coding-agent runtime and CLI for Elixir and the BEAM

README for M Shehroz Jamshaid

A Plan-and-Verify LLM Agent for Repository-Level Data Engineering

MemGuard - eBPF-based real-time memory poisoning detection and alerting for LLM agents (AgentSight overlay)

A delivery-discipline behavior layer for AI agents — turns “the agent says it's done” into “the agent proves it's done”: pre-implementation gate, one backbone + two collaboration forms, evidence-based review, real-environment acceptance. 把「Agent 说做完了」变成「Agent 证明做完了」的交付纪律行为层。

PACT: Can Enterprise AI Assistants Be Trusted Under Pressure? A benchmark of whether LLM assistants keep following compliance rules in regulated workplaces when a deadline, a manager, or a pushy user makes breaking them convenient. Paper, dataset, leaderboard, and evaluation harness.

Free AI skills and workflows for nutritionists, including guideline-to-skill automation and nutrition loop assistants; maintained by Wang Runyuan.

Autonomous Network Attack Forecasting & Multi-Agent SOC Co-Pilot powered by Temporal World Models, MITRE ATT&CK, CAPEC, CVE/NVD, and Dual Statutory Reporting (CERT-In + NCIIPC). Built for SIH 2026 (PS: 26153).

Examine consequential business and technical decisions through five distinct AI perspectives, blind peer review, and a synthesized recommendation. One model-neutral skill with native subagents for seven AI harnesses.

Python runtime for multi-agent LLM systems — commands, workers, MCP, memory, Apache SDK.

Production-grade agentic RAG system built from scratch. LangGraph multi-agent orchestration, MCP, enterprise middleware, and RAGAS evaluation, with real production bugs diagnosed and fixed along the way.

The Awen Engine — a local AGI memory organism that dreams. Recursive Harmonic Framework (RHF) reference stack: Gnostic Engine, Sovereign Client, Echo Protocol, Tesla Soul Engine.

Headless CLI automation pipeline for compiling, backtesting, and risk-ranking MQL5 Expert Advisors on MetaTrader 5.

An AI operating framework for agents and assistants to improve reasoning, decision-making, execution, and practical outcomes.

C++20 SDK for the Agent2Agent (A2A) Protocol: build A2A clients and servers with REST, JSON-RPC, gRPC/protobuf, streaming, auth hooks, CMake, vcpkg.

agent-fox is an autonomous spec-first coding agent.

弈鉴｜基于 KataGo 证据链的本地围棋 AI 私教 Agent

A multi-user load benchmark for Long-Term Memory (LTM) systems

Interactive Secure Edge AI Governance Playground with MCP, deterministic policy gates, two-person approval, attestation, and safe rollback.

run https://pi.dev in a container

Vibe coding with a custom agent ext. and a web-browser-backed agent (UWA)

Production-grade AI automation proof-of-work: structured workflows, tool/API adapters, human approval, validation, auditability and automated testing.

Agentic return-abuse risk scorer - Razorpay AI Buildathon Track 02. Live: https://return-risk-agent.onrender.com

Governed AI-assisted software work: practical control, evidence, and change-boundary tools.

Production-grade assurance framework for safe, human-centered AI: evaluation, human calibration, adaptive routing, resilience, and release gates.

HQE (/HQE) is a portable, production-grade agent skill that equips AI coding agents with the rigor, skepticism, and methodical precision of a Principal Staff Software Engineer and Principal Security Auditor.

Autonomous agentic development suite powered by Zed Editor (Rust) + OMP (Agent Client Protocol).

TUIKit is a library for helping simplify the process of making beautiful terminal user interfaces (TUIs) in C#

Catch documentation drift by safely validating explicitly marked code examples in Markdown.

Agentic SOC alert-triage service: ingests from Splunk, CrowdStrike, AWS GuardDuty and Sentinel, normalizes to OCSF, and runs a LangGraph multi-agent DAG with MITRE ATT&CK-routed enrichment to emit a structured TriageVerdict. Python, FastAPI, Redis, Langfuse, pluggable LLM backend.

Your agent stops paying to read files. Content-addressed editing for coding agents: lines, not files. ~25,000 tokens per edit, down to ~50.

Seize the means of production from our agentic overlords ✊

A local AI agent built as your personal assistant free and secured to have access over your computer and do tasks you command or chit-chat via voice or text on the chat box — Inspired by Jarvis, Chatgpt, Claude, OpenClaw, and Hermes Agent.

Intent‑aware governance gateway for autonomous AI agents that interact with enterprise APIs.

A curated guide to operating AI agents in production: observability, evaluation, tracing, guardrails, deployment, security, and incident response.

No longer updated. Early streaming engine: slower, less safe, and behind Global Agent. Current benchmarks and runtime: https://globalagent.dev

Open-source Agentic AI Academy for engineering trustworthy AI agents from learning to production: agents, RAG, MCP, evaluation, security, observability, governance, enterprise architecture and production AI.

Open-source AI copilot for Android: understands the live screen via the Accessibility API, resolves the exact UI element you need in natural language (Hindi, Tamil, English), and visually guides you to it — built for people who find modern app UIs hard to navigate.

A hands-on, from-scratch repository for learning and building multi-agent agentic AI systems. Explore Agentic-Rag, MUlti-agent RAG, ReAct, reflection, planning, human-in-the-loop, multi-agent , hierarchical , and RAG patterns using LangGraph, Python, and Jupyter notebooks.

⚒️ Opinionated agent harness TUI

Multi-model orchestration for Claude Code and Codex, powered by Antigravity with native Windows ConPTY support.

Gauntlet Mini: the gauntlet loop with a mini bill. Builders, blind critics, and a 50/75/95% usage-budget ladder so the loop converges instead of torching your limits

Nerve — an open-source visual command center for OpenClaw and Hermes agent fleets

一个可直接使用的 Agent Skill。涵盖六大生物启发模式、六大设计原则，以及可落地的操作手册。

TestDog is a desktop test case management tool: generate UI test scripts from natural language or manual recording, then replay them as-written with Playwright to verify (everyday regression costs no model tokens). Built for end-to-end regression testing of web applications.

Long-term memory for AI agents. MCP server with local embeddings, heat-based recall, and hierarchical clustering on embedded SurrealDB.

Local conversation store and read-only dashboard for Microsoft Agent Framework. Two packages, one line each, nothing leaves your machine.

MCP loader and registry written in rust

1,206 Bro. William Branham sermons in structured JSON format. Open dataset for Message developers, researchers, and worship applications.

The Awen Engine — a local AGI memory organism that dreams. Recursive Harmonic Framework (RHF) reference stack: Gnostic Engine, Sovereign Client, Echo Protocol, Tesla Soul Engine.

Agent skill that distills AI study chats into a causal-chain learning wiki — mechanical checks, a regression eval gate, evidence-driven self-iteration (RSI). DSH / Codex / Claude Code. ｜ 把 AI 学习问答沉淀为因果链学习 Wiki：机械校验 + 回归门禁 + 证据驱动的自迭代。

Chat-driven Prometheus visualization with one-click Grafana export. Self-hosted, single image, bring your own LLM.

European memory system for AI agents with focus on compression and weighted information

GH-300 live demo kit. A real Python app left intentionally immature (bugs, security gaps, low coverage) across 5 structured Copilot modules. Prompt library, Skills, AGENT.md, and a reset script included.

Photoshop-shaped layer image editor for Linux: masks, adjustments, layer styles, artboards, PSD in and out, and an AI assistant that edits with you. Rust, GTK4, made for Omarchy.

Free LLM aggregator - access multiple AI models from one interface. No API costs.

Build on your own AI agent platform in one command. Pick the modules you want and get an app of your own — running on Chatfuel's infrastructure, with the skills your coding agent needs to keep building it.

GrixChat is a comprehensive, high-performance social media and real-time messaging application. It combines the features of popular platforms like Instagram and WhatsApp into a single, mobile-first experience.

Unofficial Node.js client for DeepSeek Chat with multi-turn conversations, streaming, file attachments, thinking mode, web search, and session management.

Chat with private legal documents using local LLMs. Get cited, verifiable answers on your own hardware.

📝 Discover and use powerful prompts for LLMs to enhance your AI interactions and boost productivity in coding and creative tasks.

Educational Telegram bot built with Python and PyTelegramBotAPI, featuring a number guessing game, QR code generation, age calculation, and text-to-speech.

A verifiable memory layer for language models — facts in a graph, every claim with its source.

AI-powered banking risk and compliance assistant combining fraud detection, SHAP explainability, RAG-based policy Q&A, FastAPI, Gradio, MLflow and Docker.

JinxAS is a local voice assistant for PC designed to run with privacy, speed, and control over your hardware. Voice transcription (STT with Whisper), cognitive inference (LLM with Qwen 2.5 3B via Ollama), and semantic memory (vector RAG with FAISS) all run entirely on your machine. Voice synthesis is handled by Edge-TTS

Deep Thesis：论文写作全流程智能体，覆盖十环工作流、文献知识库、实验协作、规范引用与 DOCX 交付

安全约束 RAG 问答系统：三路检索（向量/BM25/图谱）+ 生成前禁忌拦截 + 量化评测

一个轻量级工具，一键将 GitHub Star 导出为纯净的 Markdown 笔记。 | A lightweight tool to export GitHub Stars to clean Markdown files.

Local crawler and analysis toolkit for AI/Agent technical jobs on BOSS Zhipin.

面向一级股权投资人的 AI 增强工作台。BP分析·私有知识库·投资判断沉淀·跨会话记忆。数据主权归用户。

Vectara HHEM model to reduce Hallucination in RAG models

Emotional Memory + Adaptive Personality Engine for LLMs

AI assistant for a university chemistry library: grounded RAG chat + structured extraction

DocAudit AI is an intelligent, Next.js-powered SaaS application that leverages Google Gemini 2.5 Flash to automatically audit, summarize, and extract risk factors from financial and legal PDF documents.

通过Eino+gin实现的AI agent

Python API and agentic skill for Google Gemini Notebook

🧭 AINav — 最全 AI 工具导航站，收录 590+ AI 工具，涵盖写作、绘图、编程、视频、音频、智能体、RAG、MCP 等分类，支持收藏、搜索、深浅主题切换。

非官方的AgentScope Rust重构

Self-contained RAG for consumer laptops (16 GB RAM): local ingestion, hybrid retrieval (FAISS + BM25), streaming processing, GDPR-aligned privacy. Runs 10 GB corpora offline with Llama 3.2.

Convert any skill, repo, or tool into an install-ready Hermes Agent skill. 20+ formats: Agent Skills, Claude, Codex, Cursor, MCP, plugins, prompt packs.

Qofeno agent skills — real, production-only workflows for Claude Code, Cursor, Antigravity, and other AI agents

ReplyNodes is the read-only web context API for AI agents.

Internel AI Knowledge and Decision System

Grade and converge a repo's documentation as an AI-agent context source — six-dimension star rating (completeness, correctness, freshness, linkage, consistency, economy) including entry-file token cost, plus a loop that fixes the docs until targets are met. A Claude Code skill.

An MCP tool. Connect your assistant to /mcp and every tool call goes through a broker — scoped per caller, under a credential they never hold, revocable, metered and audited.

An AI-powered agricultural assistant using LangGraph to provide crop diagnostics via computer vision, satellite imagery, weather data, and web research.