Custom agents

Teach Data Engineers How to Think AI.

Fly the web on a string — a lightweight (~7MB) Rust browser-automation MCP server for AI agents. Streamable HTTP, 21 tools, HTML→PDF, plus a Puppeteer-compatible npm facade.

Core engine behind Calibrate, a framework for evaluating AI agents: speech-to-text, text-to-speech, LLM evaluation, end-to-end simulations

inplan — a Markdown editor for human ⇄ coding-agent planning. CLI + desktop editor.

Локальный семантический граф и MCP-сервер для Markdown и Obsidian

Build reusable AI Agent Teams for OpenCode with explicit roles, multi-agent coding workflows, independent review, validation, and evidence-backed completion.

Your SvelteKit app's code health, checked before it ships. Statically, no browser — SEO, Performance, Correctness, Security, Architecture, Accessibility.

Cognitive Loop Runtime for Kubernetes

100 days, 100 sources: a daily public learning log of modern AI - LLMs, agents, RAG, fine-tuning, MLOps and more. Daily on Threads, X, and Bluesky @reksas13 & LinkedIn under #100DaysWithAI.

A video-call avatar frontend for your local LLM agent (Hermes/OpenClaw) — replaces the messenger chat and serves the agent's exact slash-command menu. Open-core engine: emotion, mood, voice, build-your-own-avatar pipeline. GPU-free · MIT.

AVE - Agentic Vulnerability Enumeration. The open behavioral classification standard for agentic AI components.

Repository-scoped memory for coding agent's decisions, pitfalls, and handoffs stored in one inspectable SQLite file, so agents continue instead of starting over.

Long-term memory for AI agents with no LLM call: an MCP server and Claude Code plugin, plus local-first context compression for LLM apps.

Use the pi coding agent (pi.dev) without leaving Neovim — π², a heavily extended fork of alex35mil/pi.nvim

Agentic social media scheduler to post on multi platforms with built-in MCP

AI-native CAD agent converting natural-language requests into validated OpenSCAD artifacts with automated geometry repair and manufacturing validation.

Turn any supported coding-agent harness into a full game-dev studio - 55 specialized AI agents and 182 skills for Godot, Unity, Unreal, Three.js, PixiJS, Phaser, and R3F.

Make your specs actually comprehensive. Pair your specs with tests to make your coding agent stop improvising bad code

Local AI voice assistant with wake word detection, screen vision, 31 tools, and Iron Man web UI. Fully offline — Whisper STT, Kokoro TTS, Ollama LLM.

A CLI for reading and editing Unity scenes, prefabs, and assets with minimal token usage oriented designs in between a live Unity Editor.

Structured market intelligence for Indian financial events, macro context, and sector impact.

Generalist agent with advanced metacognition and SKILLs.

Agent skills that turn Claude Code (or any agent) into an award-winning multi-modal creative pipeline — art, music, video, story. Panel-of-Experts debate, a deterministic state graph, and an immutable continuity block that makes creative drift structurally impossible. Prose contracts, not framework code.

A multi-agent system that generates and evaluates business ideas

Manages parallel Claude Code sessions, so you don't have to.

把数据文件（CSV/Excel/JSON/TSV）变成会讲故事的分析报告的 skill（支持 Claude Code / Codex / WorkBuddy 三家编程智能体）：先跟你确认行业与报告受众，再清洗分析，产出单文件 HTML 交互报告——结论先行的六段式、每张图配解读、数据处理过程可追溯；不杜撰行业基准、相关不当因果、小样本不硬下结论

Equip the right context for the job — named context kits for AI coding agents, auto-equipped per stack, machine, or task

Chaining AI & API agents to streamline software development and achieve goals collaboratively.

Offline-first AI memory for coding — 29 MCP tools, temporal knowledge graph, and local vector search. No API keys, no cloud, no server.

A native macOS terminal for running AI coding agents in parallel — built on real libghostty, with persistent tmux-style sessions and a scriptable Lua/CLI control plane.

187 plug-and-play AI agent personas dalam Bahasa Indonesia — Indonesian translation of agency-agents + 3 Indonesian-market originals (Tokopedia/Gojek/WhatsApp Business ID)

Open source skills for GTM

源码级拆解 DeepSeek Harness · 面向 Agent 开发者的中文学习资料

Универсальные безопасные Agent Skills для любых конфигураций 1С / Universal safe Agent Skills for any 1C:Enterprise configuration

Deterministic, local-only audit reports for Claude Code AI agent sessions. Rust + SQLite. Zero network calls. MCP server for agents.

Skill-Atlas

Continual self-evolution plugin for DeepSeek Harness: versioned, auditable, rollback-safe harness state refined from session trajectories, with a benchmark-driven validation loop.

DeepSeek Harness URL reader: fetch any page and return clean main-content text/Markdown. Auto charset (GBK/GB2312/UTF-8/Big5), token-efficient (6000-char cap, cache, offset), zero deps, no API key. 网页一键读全文 → 干净正文 / 结构化 Markdown

AI-agent skill for building Lark (Feishu) slides via the lark-cli: 51-page template library, brand design tokens, XML generation workflow, and automated layout review. Built for agents, reusable anywhere.

Self-host Supabase in production, AI-ready: read-only MCP for your coding agent with zero secret exposure, SSO, PITR backups, monitoring, disk encryption, and one-command migration off Supabase Cloud.

Secure MCP access to local repositories — without shell, Git, or arbitrary writes.

Ask AI, take a note, or set a reminder right from your Mac's notch. Native & Design-first.

AIX — AI eXchange Format. A strict superset of Google's Open Knowledge Format (OKF) v0.2: stable identity, typed relationships, provenance, media identity, federation. Every AIX bundle is a valid OKF bundle.

Local-first personal workspace: write notes, track tasks, chat with an AI that knows your files - all plain Markdown, no cloud.

Your AI coding harness, ported to every client you use. Capture Claude Code (or Codex) rules, hooks, skills, subagents, commands and MCP servers once; apply them identically to Codex, Gemini CLI, Cursor and 16 more. Zero-dependency Node.

基于 Codex 的网络小说创作工具集skill — 从构思到完本的完整工作流，面向番茄、起点、七猫、刺猬猫小说等阅读平台。

Windows + Docker + Python 基于 Milvus + MySQL + Redis + 大模型的私有化 RAG 问答系统

A Model Context Protocol server that drives Aseprite headlessly and measures what it drew: 156 tools to draw, shade and animate pixel art, and results that say what each call did. An agent that cannot see the screen can diff two frames, score palette conformance, and verify a loop. Exports PNG/GIF/sprite sheets, Godot and Minecraft packs.

A self-hosted, mobile-first web GUI for Herdr with browser terminals, workspace and worktree management, file and diff viewers, and AI agent session inspection.

The Mozaiks core is a self-hostable runtime for deploying AI-driven, multi-tenant web apps with real-time UI, auth, and billing built in. Apps are defined via YAML and lightweight Python/JS stubs for agent logic, workflows, and integrations — so developers focus on behavior, not plumbing. :link: https://docs.mozaiks.ai/ :link: https://mozaiks.ai

Guarded, recoverable WordPress MCP automation for Elementor, Gutenberg, WooCommerce, content, and code - with backups, validation, and audit trails.

DSH CyberWorkStation — one-click DeepSeek Harness (dsh) bundle: cyberpunk desktop launcher, SillyTavern-grade Control Deck, 8 production plugins, AI skin studio & markets. Vendored core rc.8, run setup.cmd and go. 一键部署的 dsh 赛博工作站整合包。

Reusable skills for AI coding agents working with PlayCanvas products

Terminal-native frontend for DeepSeek Harness with real scrollback; a direct, in-process consumer of Harness capabilities.

A SKILL.md for AI coding agents that stops overengineering before it starts. Turns KISS, YAGNI, and DRY into a real decision system — not just a reminder.

Agent Skill that turns your coding agent into a keyboard firmware expert. Audits ZMK/QMK keymaps, tunes home row mods, makes trackballs layer-aware, generates keymap diagrams with CI, adds alt layouts (Graphite, Colemak-DH), and debugs the failures the docs don't cover. Sync keymaps! Works with Claude Code, Codex, Cursor, Gemini CLI and more.

A Next.js multi-provider AI workspace for chat, search, reasoning, image and video generation, and browser voice interaction.

DeepSeek Harness Web GUI 的网页预览与元素批注插件，让 AI 根据可视化反馈直接修改前端源码。

Python + Terraform. Cloud-ready GraphRAG solution that uncovers distant connections across data and grows incrementally as new documents arrive.

Fully customizable Advisor and Executor flow plugin for the Pi Coding Agent

Bring Apple Health to any AI agent—privately. Open-source iPhone app on the App Store, automatic sync to a self-hosted receiver, durable local history, and a read-only MCP server. No hosted intermediary required.

Claude Code skills for microcontrollers (MCU): STM32, ESP32, RP2040, nRF52840, AVR board-specific Agent Skills (SKILL.md) for embedded firmware

Local document knowledge base for DeepSeek Harness: library_add/remove/list, hybrid semantic+keyword library_search with diversity re-ranking, relevance filtering and lost-in-the-middle avoidance, citation-aware injection, library_cite_check and library_diagnose — SQLite-backed index via the storage domain, local embedding, zero model downloads.

Build production marketing sites with Claude Code — an Astro + Cloudflare Workers template, a WordPress→Astro migration playbook, and 36 silent failures documented from real rebuilds.

Jump Skill Ninjas are intelligent meta-skills — master orchestrators that route your requests to the most appropriate specialized skill from a curated library. Instead of manually searching through hundreds of individual skills, you simply describe your task and the Ninja Identifies specialized skills and Loads the full skill in context

Self-hosted A-share multi-agent research desk: 15-agent debate, optional Tushare Level-2 order queue, K-line workstation with AI insight, and a signal translation layer so models read conclusions — not raw dumps. AGPL-3.0.

Parent-operated starter kit for designing a private, child-facing Hermes Agent profile

Open-source MCP control plane for AI coding agents: mechanical time-locks, evidence-gated proof receipts, red-team remediation, and persistent engineering memory.

The open-source safety and control layer for personal AI agents: Engine, Desktop, Mobile, proxy, and shared contracts.

纯 Kotlin 原生 Android AI 智能体：多模型接入、子智能体协作、在线插件市场、技能自动引用、root/Ubuntu 终端、长期记忆

Goal. Loop. Audit. Done. A pi-coding-agent extension that supervises long-running work, with an isolated auditor on each completion. Beat bamboozling by design: the auditor runs in a fresh session with no extensions, no skills, no editor, only the read tools needed to verify your goal.

Evidence-gated Claude and Codex council orchestration.

2 text files and a folder. The ICM Folder System I run several businesses on, in plain English, for people who do not write code.

Reference harness for the Big Finance benchmark of workflow-grounded financial-research questions

Official OpusClip skill for AI coding agents — clip long videos and post to social, from Claude Code, Codex, OpenClaw, or Claude.ai. Bundled CLI, bring your own API key.

Use [Multi]ple AI agents for code-review, verifying and questions

Remote Desktop Control for AI agents: screenshot, mouse, keyboard and window control of another machine over Tailscale, exposed as MCP tools. Built with Rust

A portable AI agent skill and architecture compiler for polished, editable, evidence-grounded Draw.io diagrams.

A searchable, source-linked atlas of AI coding-agent commands, flags, shortcuts, MCP, permissions, skills, and sessions across Codex, Claude Code, Gemini CLI, Cursor, Copilot CLI, and Muse Code.

RoboClaw live robot lab — community, experience, and API. Star to unlock the Host Concierge.

Research Software Engineering skills for AI coding agents

A fast, provider-independent coding agent for the terminal, written in Rust.

Open-source, local decision models inspired by TypeSafe’s Jev and System One - built on Qwen3 for NVIDIA DGX Spark.

Unofficial MCP server for AI-driven NI Multisim circuit generation, simulation, data export, and reports

Agent skills for frontend UX/UI design that actually looks good

三菱 PLC 编程与工程助手。 MCP server and AI workbench for Mitsubishi PLC programming with GX Works2. Generate Ladder and Structured Text, edit native FBD, and review versioned projects with CPU-aware RAG and diagnostics.

A Flutter client for herdr: a status board and live terminal for the coding agents on your own machines. Over SSH, with no bridge binary to install.

Validate SKILL.md files, track agent-skill bundle changes, and check release evidence against trusted policies.

AI-powered career workspace for job tracking, resume management, and interview preparation.

A state-based autonomous agent framework. Crafting digital entities with non-linear cognitive memory and self-evolving emotions. (Work in Progress)

Reproducing agentic KV-cache policy claims on real traces. 68k requests from 393 Claude Code sessions. LRU is harder to beat than the papers suggest.

MCP server for managing Gologin browser profiles, proxies, fingerprints and cloud browser sessions through AI conversations

Six skills tell coding agents how to work. Two guards enforce the parts that should not depend on judgment.

Профессиональные Data Engineering Agent Skills для разработки AI Agentic Data Platform

Every GTM engineering project worth building — built live, in public, with Claude Code. Companion repo to The New GTM Engineer.

Acta: A Decision-Centric Runtime for Agents

Real-time drone swarm autonomy simulation using Jev for fast System 1 reflex decisions and collision avoidance, with optional System 2 reasoning for strategic guidance

AI agent skill (Claude Code / Codex) that turns your product into a cinematic promo video: 152 shot recipe cards, 209 style variants, tuned Remotion implementations, 149 SFX, a validated 10-shot template, plus music, voiceover, captions and B-roll via the dLazy CLI.

Give an AI agent a spare Android phone: self-hosted MCP, companion APK, live web control, and Chinese app walkthroughs.