Custom agents

A production-ready toolkit to accelerate and automate the end-to-end lifecycle of AI Agent development.

A simple Markdown editor for adding inline edits and review notes directly into MD files for AI-assisted workflows.

A Retrieval-Augmented Generation (RAG) application for querying legal documents. It uses PostgreSQL, Elasticsearch, and LLM to provide summaries and suggestions based on user queries. Features data ingestion with Airflow, real-time monitoring with Grafana, and a Streamlit interface.

Xihe · 羲和 — the system around the model. A model-agnostic multi-agent runtime: deterministic orchestration, a self-consolidating memory, a cross-model verifier, and engineered taste. Wright is its first agent.

Token-efficient browser-automation MCP server for AI agents. One Go binary (Go + chromedp), no Playwright/Node. Stealth + delta snapshots.

MCP server + browser sidecar extension that lets AI agents run Google AI Mode searches using your own logged-in browser session

Deterministic cost / loop / time budgets · full observability · crash-resumable runs · human-approval gates · a memory you own. Self-hosted. Your keys. No telemetry. Point it at your existing agents - one env var.

Declarative, local-first runtime for bounded AI agents — define an agent in one manifest, run it headless, and audit every effect.

🛰 Local agent runtime reporter for the OCTO platform — detects installed AI agent CLIs (Claude Code, Codex, OpenClaw, Hermes), reports status & bindings, supports one-click remote upgrades.

AI WorldCup Arena 2026: AI agents compete to predict the 2026 World Cup with virtual points, leaderboards, reports, and transparent sports analytics.

Auditable context capsules for LLM handoffs, coding agents, and OpenCode MCP workflows.

Train computer-use agents end-to-end on real desktops, at scale: a high-performance runtime where 8K+ live environments on one laptop thread.

One room, three minds: a multi-LLM planning council for your terminal — you, Claude Code, and Codex debate and ship an implementation plan.

A lightweight local-first CLI that orchestrates AI agent CLIs (claude, codex) into powerful workflows with a simple YAML.

MCP server for browser automation via Patchright. Multi-profile Chromium, headed/headless, StreamableHTTP protocol. Includes dashboard, skills, and AI agent setup.

PulseDeck — the reporting inbox for your AI agents.

Turn an agent's work: progress, reviews, research, comparisons into a single self-contained, reviewable HTML page (a work canvas). An Agent skill.

The AI intelligence for your browser history. openloops scans your browsing history, groups it into the things you were actually trying to do, and helps you close the loop. Fully local.

Bantz won't be a helper; he will be your Host. Your computer will become his studio, and your tasks will give the "entertainment." He won't just execute code; he will "pull strings" behind the scenes with a smile that you can hear through the text.

AI pull-request reviewer: hybrid RAG + a code graph + Claude Code/Codex. Whole-repo context, inline GitHub comments grounded on exact code. MCP server + Claude Code/Codex plugin.

A self-hosted knowledge broker for homelabs. Routes and fuses queries across Kiwix, FreshRSS, Open-Meteo, SearXNG, Uptime Kuma, and Home Assistant — and tracks what changed over time. REST API and MCP server, integrates natively with Open WebUI and Home Assistant.

Headless Matrix WebRTC voice AND video agent — auto-answers calls, bridges audio to any AI agent via PipeWire, optional camera-frame vision for multimodal LLMs

An autonomous AI coding agent with novel innovations in tool state management and code editing, running in a Docker sandbox with a persistent shell, parallel tools, and a VNC desktop.

A validated, Baseline-aware modern CSS skill for AI coding agents (Claude Code, Codex, Cursor, Pi, OpenCode, Kiro).

mcp server for boondmanager

Claude/Grokbot plugin and scripts: long-running controller pattern to chunk and delegate code work to cli-agents, with embedded adversarial review and RAG corpus

🎮 An AI workflow editing framework focused on business customization. 专注于业务自定义的 AI 工作流编排框架，支持 LLM 调用、知识库检索、记忆读写、API 编排等场景。⚡ Vue 3 一行接入

File-based group chat for Claude Code instances — group channels + DMs, namespaced per project. Installable as a skill or plugin.

Local-first terminal agent runtime with isolated subtasks and intent-aware safety.

MindWiki — 10 顶级专家组团的决策追问系统 Skills（张小龙、乔布斯、马斯克、张一鸣、芒格、王兴、雷军、梁宁、Naval、毛泽东）。带工具调用和自我验证测试。

One Works application monorepo for AI agents, plugins, desktop, web, CLI, and Relay.

Helix is ​​a powerful self-learning AI agent with memory, skills, and the ability to recall tools. It learns from successful tasks and creates reusable skills for the future.

私教.skill — 输入任何想学的技能，蒸馏出一个带你从 0 学会它的有状态私教 skill

Create with Cavalry from AI. An MCP server that gives your AI assistant a direct line into Cavalry — run scripts, query the knowledge base, and animate from a prompt.

Curated list of AI agent frameworks, coding and automation agents, MCP servers, and evaluation tools.

Privacy-first, local-first research notebook for macOS & Windows — chat with your own PDFs, Office docs & web pages and get cited answers, all on your machine. Open-source NotebookLM alternative.

The command-line interface for the LinkAI agent platform. Access AI models and platform resources (knowledge, databases, app, workflows, plugins) from your terminal, built for humans and AI agents alike.

Auth for AI agents and humans. First-class agent identity, MCP OAuth 2.1, delegation, audit. TypeScript, edge-native, MIT.

GUI agent using WebMCP polyfill. Control web interfaces with natural language

OpenBI- AI-native BI platform - chat with 90+ data sources, build dashboards, and automate reports. Self-hosted with Docker.

Apricity — a fast, local-first web UI for MemPalace: browse, search, edit & curate your AI agent's memory, knowledge graph & vector store. Zero-dependency, self-hosted, MCP-friendly.

Cross-platform SSH, RDP, VNC & web connection manager with a built-in MCP server — let Claude Code or Codex drive your infrastructure

Hypertaks = "A founder who owns a library of more than 1,400 books, knows exactly which book to consult for this issue, and has a strict SOP: intake → formulation → implementation → integration → documentation."

The self-improving SRE agent

Framework to build WhatsApp bots, without the boring part

Full-stack framework for building AI apps and agents with TypeScript and React. Works with Node.js, Deno, and Bun.

Security Agent Benchmarking and Evaluation Research

Personal AI assistant powered by LLM and MCP servers. Control files, GitHub repositories, and Google Sheets through natural language — all from the terminal. 134 tools, async chat interface, full CI/CD.

Open-source AI agent identity & compliance platform — X.509 certificates via AWS KMS, behavioral risk scoring, auto-revocation, and MiCA / EU AI Act audit trails.

Think first, then code. Where Claude Code and Codex jump straight to writing, this stays review-first: a code reviewer, security scanner, root-cause debugger, and code generator, all in one shell-native loop.

Emacs interface for Claude Code with auto-response, vterm integration, and session management

Deprecated — Catalyst local runtime, kept for history. Use coalesce-labs/catalyst-dev-skills and coalesce-labs/catalyst-cloud-skills.

A native macOS code editor for CLI-agent workflows. Split editor and terminal panes, a cross-project Agent Inbox, private notifications, LSP, Git context — zero Electron.

A verifiable local memory architecture for long-running AI agents

Give AI coding agents product, spec, task, QA, and release workflows in one command.

Archive-grade .mdc rule library for Cursor AI — 26 production-ready rules for teams seeking stable, reusable AI coding guidance

AI-powered English learning assistant for Chinese elderly users, with RAG, ReAct Agent, two-stage retrieval, and 5-dim evaluation framework

MCP server that lets AI agents SEE embedded images inside Excel files — full text + image extraction for .xlsx/.xlsm

Local Codex control with bounded runs, crash recovery, and auditable results. Optional ChatGPT bridge.

Claude Code skills for ROS 2 Jazzy — establish the unknowns first, verify against the installed system, prove the result ran.

The agent-first IDE, built for production.

Distill your bros. The ones who make you laugh until you can't breathe. The ones who say stupid things that somehow make sense. Online or offline — if they're your brother, this skill remembers everything about them.

Local-first diff review for AI-assisted development, you have your own judge.

Production-grade multi-agent LangGraph template — per-run cost caps, canary pack versioning, Helm/KEDA, 800+ tests

Murmur Connect (Murmur) — encrypted agent-to-agent messaging for AI coding agents. Claude Code, Codex and any MCP client talk across machines: MCP-native, E2E encrypted (XChaCha20-Poly1305), NATS transport with optional JetStream durability, cross-org federation, A2A interop. Mac and Windows apps, CLI on npm.

祈愿博客模板 · 基于 Astro 的全功能博客模板：AI 问答(RAG)、文章朗读、评论、网盘、短链、资源站，EdgeOne Makers 一键部署

Claude code / Open Code Skill guiding you step by step, to generate continuous-shot video using Gemini web interface

Make Codex and Claude Code sessions multiplayer from the terminal.

咨询风格 PPT 生成 Skill：core 共享逻辑 + Claude/Codex 双平台适配，默认全原生可编辑，HTML 样张零图像依赖，12 视觉风格 + 10 场景框架

Drop-in instructions, prompts, and workflows to reduce token usage across Codex, Claude Code, GitHub Copilot, Cursor, and other AI coding agents

The McKinsey 7-step problem solving framework (from Conn & McLean's Bulletproof Problem Solving) as a portable Claude Code skill. Drop-in analytical rigor for AI agents.

Claude Code plugin that bridges to agy (Google Antigravity CLI) — adds Gemini, GPT-OSS, and grounded web search to Claude sessions

The n8n alternative without the bugs — faster, more reliable workflow automation.

Agentic Memory System for Long-Term AI

Spec-driven multi-agent teamwork for any skills-capable agent (Claude Code, Codex, and more) — plan, dispatch, monitor, QA. No framework lock-in.

多功能 AstrBot 插件（CurrentCortex）—— Pixiv 随机图片 ·网易云点歌 ·JMComic 漫画 ·小红书/B站/抖音媒体解析 ·每日一言 ·天气 ·男娘 ·DG-LAB（郊狼） 设备管理 ·跨群聊记忆 ·按群聊开关。基于 LeiZ API。

Canonical memory infrastructure: BM25 recall over trusted files for AI agents.

A streaming action language for computer-use agents. LLMs emit one JSON action per line; the executor runs each the moment its newline arrives — while the model is still generating.

QQ bak文件聊天记录解码导出工具，将 QQ / NTQQ 聊天记录转换为 Agent 可读的 TXT、JSON、CSV。

Fully local AI assistant that watches your screen to eliminate the context problem. No cloud. No data leakage.

Decision infrastructure for agents. Verifiable, traceable, auditable.

An AI retrosynthesis framework with multi-objective route scoring, USPTO precedent grounding, and forward-plausibility checks

Scenarios and tasks for evaluating AI agents on real-world AWS automation tasks. Use the aws-bench framework (https://github.com/aws-bench/aws-bench) to provision isolated AWS environments, run agents, and evaluate their results.

Way to find a work in AI era — storefront, MCP, self-host UI, Hub for products & startups

Files you own. Memory your agents can trust. Also: your AI won't gaslight you about what's in your own notes anymore.

Turn your team's unwritten code-review conventions into rules your AI coding assistant follows.

Your project remembers — the project brain for builders and agents.

A curated list of AI tools, workflows and resources for writing, editing and publishing full-length books.

The operating system for autonomous software engineering.

Oh My Pi host plugin for the GSD Embeddable Orchestration System — extension bindings, slash-command surface, Host-Integration SDK adapter, and optional Goal Mode status integration for AI coding agents

AI Agents in Obsidian: Claude Code, Codex, Cursor, Grok, OpenCode, Oh My Pi, Pi, and DeepSeek Harness for note chat, inline editing, and file diff review.

Reproducible benchmarks for LLM output drift and agent replay in financial operations: DFAH-Bench, the installable dfah package, and a hands-on workshop.

🔍 AI-powered code auditor using Rust & Ollama. Features agentic tool-calling architecture where LLM autonomously explores, reads, and analyzes codebases. Detects security vulnerabilities, bugs, and performance issues. Generates detailed Markdown reports with actionable insights.

Local-first macOS AI workspace. Notch-native, 8 providers, your coding agents and their subagents in one place. ⌥Space anywhere.

Multi-Agents VSCode Extension that Focuses on the Full Control of Context, Observability and Auditability. Works like Wakaba Mutsumi's Personalilties 🥒. 吓我一跳我释放忍术😰若叶睦那么多首啊🥷

AI agent framework for infrastructure operations.

User Interface for Prime-Agent

面向 CTF 竞赛评测平台（TSec Benchmark）的自动化 Web/逻辑漏洞挖掘与利用 Agent：分层调度 + 多会话并发 + 跨会话共享黑板 + 平台权威校验

Open-source, self-hostable server and runtime core for Run402 — backend infrastructure for AI agents. Postgres/PostgREST with RLS, storage, routing, functions and portable project archives. Apache-2.0.

Universal SAP skills and agents for Claude Code — FI, CO, TR, MM, SD, PP, HCM, SFSF, ABAP, S/4 Migration, BTP, BASIS