Custom agents

A bilingual (EN/ZH) 4-stage research workflow skill for Claude Code and Codex CLI — topic refinement, urgency assessment, route evaluation, and experiment planning with citation verification

数学建模论文生成器 (Claude Code Skill)：给一道竞赛题，按国赛标准四阶段流程（策略→求解→写作→审查）端到端生成可编译的 LaTeX 论文。LaTeX-only 分支，xelatex 编译，MATLAB MCP/Python 跑真实数据，自带国赛标准模板。

Get insights into token usage for Claude code and Codex sessions with a dashboard and menu bar widget for mac

Framework de governança para desenvolvimento assistido por IA (Vibe Coding). / Governance framework for AI-assisted software development (Vibe Coding).

MCP vision server for AI agents, Claude, Cursor, OpenCode, VS Code, local LLMs, Ollama, LM Studio, OpenRouter, Gemini, image analysis, and video understanding.

YAML hook automation for the PI coding agent: tool guards, session hooks, prompts, notifications, and bash actions.

为 DeepSeek V4.0 等单模态大模型装上眼睛和耳朵 —— 本地视觉+音频 MCP 服务器，基于 Ollama + MiniCPM-V 4.6 + faster-whisper，图片描述·视频分析·语音转文字

Agent-agnostic semantic skill for MaxCompute SQL — CLI + bare SKILL.md bundle

Professional, evidence-based reviews. A portable skill for Codex, Claude Code, Gemini CLI, and other AI assistants. 好用的 paper review skill，经过专业研究人员长期使用演化而来。

Harness engineering for Claude Code, run daily: hooks, skills, typed memory, roles, a delegation queue and self-audits that let you hand a coding agent work and trust it while you're away. 18 load-bearing patterns, an interactive tour, a learn track, teardowns of real systems, a CI linter and fork-ready samples.

A compact, SDK-only agent harness: maximum performance, minimum framework code

Desktop provider switcher for Claude Code and Codex. Keep every endpoint, API key, and model in one place, switch the active provider in one click, and probe latency before you start. Atomic writes and rolling backups keep your configs safe. Python + Next.js in a native OS webview.

30 MCP tools to drive the Godot Editor from any MCP-compatible AI agent — scenes, nodes, debugging, and screenshots, in both Editor and Play mode. OpenClaw gateway included for remote and chat access.

Local SQLite memory for LangGraph agents, with provenance, TTL, export, and MCP.

List of all the awesome projects into ADK-Hackathon

The project's aim is to develop a chatbot that can help people create spec documents without knowing the specification.To get started with, the bot will consume the spec, JSON schema and serves the user as an expert. So based on a set of questions and answers it will generate an AsyncApi spec document according to the use cases.

AI agent memory you can trust

Pier Code is a terminal-native agentic coding tool that understands your codebase and helps you move faster by handling routine tasks, explaining complex code, and managing git workflows through simple natural language commands.

An end-to-end framework for running, sandboxing, and scoring agentic LLMs on complex data-science and econometric replication tasks.

Alpha-stage PRD generation with FastAPI, Streamlit, provider adapters, streamed state, and explicit scope boundaries.

Multi-agent AI Scrum team for Claude Code — native Mac app + CLI. Autonomous Sprints; you stay Product Owner.

Cross-agent coordination for AI coding agents. One daemon, one database — so multiple agents in one repo can see each other's plan, tasks, and file claims.

A browser for LLMs, not humans.

Snapshot testing for LLM apps and agents, built to run locally and block regressions in CI.

高性能，低门槛的通用终端 AI 代理工具 / General-purpose terminal AI agent

AI smart home assistant for Home Assistant: LangGraph agent + MCP tools + RAG knowledge graph + vision perception

It includes everything in AI like Tokenization, How does LLM works, embedding, Roadmaps, Andrej Karpathy video sum up, and much more.

Tell Claude Code what you want once. It works until the job is verifiably done — and gets smarter every time. Self-correcting goal loops with independent verification and growing memory.

A Selenium-based chatbot for Kick (Kick.com) that automatically logs in using stored cookies, sends messages from a text file to a specified channel, and supports both headed and headless modes. Features include sequential message sending, status indicators, real-time message display, and adjustable message intervals.

Make the AI coding agents on your machine (Claude Code, Codex, Reasonix, ZCode) work as one team. Zero-dependency, local, MCP + CLI.

⚡ Multi-Device WhatsApp Bot Engine with Hybrid OpenRouter AI, Modular ESM Plugins, RPG, & E-Commerce Automation

CLI + MCP for agent context: a task layer over TickTick, Notion, GitHub, Obsidian & more (RRF retrieval, provenance, reviewed writes, undo) plus a knowledge-graph layer — ats kg: facts from any source, human-ratified. Recommended engines: Graphiti (server), LadybugDB (embedded).

Simplified Technical English for Code - extract, adapt, and enforce ASD-STE100 rules for software documentation. 9-agent pipeline, 59-test benchmark (96.6%), multi-language scaffolding.

AI-orchestrated, Kali-native pentesting lab: an MCP tool-server with scope/VPN/rate/audit enforced in code, independent finding verification, MITRE ATT&CK + OWASP mapping, a reusable findings library, retest tracking, WireGuard OPSEC, LUKS encryption, and automated PDF reports.

Graal Script Agent is a Java library for adding AI-authored extensions to applications.

Eval-driven multimodal agent for identifying trees, logs, bark, leaves and wood from photographs. Dendrology is the reference domain; the engineering subject is evidence handling, calibrated uncertainty and evaluation.

轻量高颜值AI化的跨平台SSH客户端，支持SFTP、端口转发、串口

把任意内容自动做成设计师级视频的端到端 skill:AI 写分镜 JSON,Remotion 渲染。8 步流水线:读取→定调→方向→分镜→配音→分段渲染→合成体检→导出。

내 맥에 상주하는 AI 팀. 텔레그램·슬랙에서 일을 맡기면 누가 담당인지 정하고, 진행이 대시보드에 남습니다.

Local-first agent that drives Claude Code and Codex over ACP. Desktop app, git-URL plugins, A2A 1.0. MIT.

Kimi Code-style agent orchestration harness for the Pi coding agent — Swarm, Goal, Plan, Permission, Task, Hooks, Skills, TUI

Add an agent with one CLI command. Source lands in your repo as real TypeScript — prompts, tools, and runtime you can edit and own.

Claude Code에 '15명 전문가 관점 + 자동 트리거' AI 개발 파트너 페르소나를 입히는 플러그인. 항상켜짐 hook + 조건부 skill + 인터뷰형 생성/편집 명령어, 세무·법률 면책 자동, 새 PC 이식형. Apache-2.0.

A 125-task benchmark for knowledge-mediated implicit associations and retrieval blind spots in long-term agent memory.

Zero-token memory for LLM agents. Rust implementation of Zero-Mem (arXiv:2607.29377) with a Hermes Agent memory provider.

Multi-tenant AI customer support platform, businesses upload docs, FAQs, PDFs, or a website URL and get a RAG-powered assistant that answers customers 24/7.

Skill-guided DevOps agent built on MECE for chat, shell, VPS deployment, Docker, GitLab, AWS EC2

AI reads HN 4x daily, writes spicy digests. t.me/claudehn

A personal substrate for agent packages that ship their own screens. One manifest declares the agent, its verbs, and its UI.

🦞 Headless real-time Python voice assistant powered by ElevenLabs, Gemini, Mem0, MCP tools, and local wake-word detection.

A 7-module course for building your own AI Chief of Staff in Claude Projects - no coding required

Regression testing and configuration sweeps for RAG pipelines, with the statistics to know whether a change actually helped.

Automatically loads the best free OpenRouter models into Hermes Agent. Ranks every :free model daily by privacy (optional) and rotates your default and fallback models before they expire so you can set and forget.

Evidence-backed proof of done for Claude Code, Codex, Cursor, and other AI coding agents. Run real checks, bind results to current Git state, and know what is ready for human review.

Cheap, isolated copy-on-write git worktrees for running many AI agents on one repo — an extra checkout costs metadata (~0.4 KiB per tracked path), not another copy of the tree. No daemon: just git + reflink/overlayfs.

Unofficial Windows desktop client for the unmodified DeepSeek Harness Web UI

Give life to an AI agent with a single command.

mahmud-fca is a Facebook Chat API modified with bug fixes, stability improvements, and extra features for Facebook Messenger bots.

20+ reference plugins for Ouroboros, the Agent OS — autoresearch, github-pr-ops, guardrails-eval, langfuse-observability and more, plus the UserLevel plugin contract.

Turns natural-language requirements into executable Python software, tests it in isolation, and packages only independently validated builds.

Multi-model, multi-agent OMP development workflow with Graphify

Installable Tauri desktop distribution for DeepSeek Harness with a bundled runtime, trusted plugin review, and daily compatibility checks.

Curated open-source small tools and workflow recipes for Chinese e-commerce/retail: Douyin/Xiaohongshu/WeChat Channels scraping, Taobao/1688/PDD/JD ops, WeChat SCRM, live-streaming, logistics, Chinese NLP, cross-border. Universal tools kept and tagged.

Experimental DSH Minimal trajectory-steering adapters for DeepSeek V4 Pro on Pi, Oh My Pi, and OpenCode V2. 推理轨迹引导实验。

⚡ Next-Gen Modular WhatsApp Multi-Device Bot Engine built with Node.js (ESM), Smart AI Assistants, RPG Ecosystem & Digital Store Automation.

A curated list of DeepSeek Harness (DSH) plugins, desktop clients, marketplaces & tutorials · DeepSeek Harness 开源项目与教程精选（中英双语）

Odyssey is an open-source, bundle-first agent runtime written in Rust

Open-source agentic AI operating layer for hospital operations — multi-agent orchestration over FHIR-connected HIS data, built on LangGraph and Temporal.

Playwright browser automation for DeepSeek Harness｜面向 DeepSeek Harness 的 Playwright 浏览器自动化插件

Generate publication-quality DESIGN.md files from any website URL. Extract computed styles, structured tokens, and layout patterns into a unified reference for AI UI coding agents.

Cyberpunk multi-agent command center for DeepSeek Harness — Chinese-first conversation, code execution, VS debate and trajectory workbenches.

Discovery-first cold outreach for solo founders — find the companies other tools miss, draft each email in your voice, send from your own Gmail. Local, single Rust binary. Built on Canonical.

Local-first AI character companion framework: emotion × energy × affinity state machines, built-in memory pipeline (facts, vectors, day summaries), background reflection, and proactive speech after silence—architecture-first “persona bones,” not prompt-only chat. Electron/Vue + FastAPI; optional VLM & TTS. AGPL-3.0.

Local-first, LLM-agnostic memory system — hybrid retrieval (vector + lexical + graph), multi-signal ranking. Runs on 4GB RAM, CPU-only, no cloud.

Mail for AI agents ( and humans). MCP server giving Claude (or any agent) safe, permission-controlled access to email (Microsoft Graph + IMAP), calendar and your knowledge files. Two-phase confirmation for sends, audit log, local index. No built-in LLM: your agent brings the intelligence.

Hosted Google Search / SERP MCP server from HasData, no Google Cloud project. Organic results, AI Overview, People Also Ask, AI Mode, news and shopping for Claude, Cursor and any MCP client.

문서만 주면 데이터 생성·학습·평가까지 알아서 도메인 LLM을 만들어주는 Agentic LLMOps 에이전트 — 자연어 지시 + Human-in-the-Loop, 로컬 구축. Chat-driven LLMOps agent that builds domain-specific LLMs from your documents.

A curated distribution of DeepSeek Harness. Overlay, not a fork.

a Lua-scriptable experimental coding harness

Unofficial JetBrains plugin: adds a Claude Subscription (Pro/Max) ACP agent to the AI Assistant chat agent picker

Send messages from AI agents, CLI, and MCP tools - powered by one shared TypeScript core.

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.

面向个人与 B 端销售的本地客户军师、成交教练与混合 RAG 决策助手

Measure whether an embedding upgrade is worth it, bridge it without reindexing when it is, and migrate safely when you are ready.

MCP server for ICTFax — list and track faxes, upload documents and send faxes over ICTCore REST. By ICT Innovations.

MCP server for ICTExam — read exams, gradebooks and item analysis; parse papers with AI and publish exams. By ICT Innovations.

Entente — provenance engineering for coding agents: handoffs become confirmable, verifiable, retryable contracts; agents form a network with evidence and a way back

Trusted AI mathematics workbench: ProblemContract to Attempt to verified proof/counterexample bundles with SymPy and Lean evidence gates.

Measurement for coding agents building UI: extract a contract from the reference page, check the implementation, get numbers back

Hybrid vector + knowledge graph RAG system for enterprise documents. Fuses Pinecone dense vectors, BM25 sparse reranking, and Neo4j multi-hop graph traversal with agentic query routing, cognitive multi-tier memory, and grounding audits

DeepSeek Harness (DSH) 跨会话记忆插件：对话模型边干边记的 Markdown 项目记忆，支持专题文件渐进式披露、可选后台静默蒸馏、按项目隔离与热更新设置面板。机制对齐 Claude Code auto memory。 | Agent-maintained cross-session Markdown memory: progressive-disclosure topic files, opt-in silent distillation, per-project isolation. Claude Code-aligned.

Agent Skill: plan-driven software engineering at scale — versioned, ordered, dependency-aware sprint/task plans with build+test gated execution.

A production-oriented multi-agent AI Agent system for software engineering. | 面向软件研发任务的多智能体 AI Agent 工程系统

Agent skills for ideas or uploaded images to printable 3D sculptures: Hunyuan3D drafts, mesh checks, slicing and verified print handoff.

A Codex CLI plugin packaging a full software studio — 45 specialist roles, 91 workflow skills, and deterministic quality gates for game and app/web/service development.

Agent Linux for AdaptixC2

Fast browser agent for ego lite. One TypeSafe request per step; an agent or Jev picks the move.

A browser extension that drives local Agent CLIs that speak ACP from within your browser for web information gathering, web automation, and more. 一个浏览器扩展程序，支持在浏览器中驱动本地 AI Agent 进行网页信息采集、网页自动化操作等工作。

An agent lives in your folder, a wiki grows around your files. Browser-only, local-first AI knowledge base — your key, your files, your machine.

LuckyAgent — 注重长任务规划和系统记忆性的Agent框架