Custom agents

Build, deploy, and orchestrate teams of Claude Code AI agents on Slack

Korean translation of "Harness Engineering: From Claude Code Internals to AI Coding Best Practices" — Claude Code 소스 코드를 체계적으로 분석하여 AI Agent 시스템의 핵심 엔지니어링 패턴을 추출한 기술서의 한국어

The framework for agent-driven websites

An autonomous, 100% local Agentic GraphRAG system for codebase intelligence. Paste any GitHub URL or local directory to chat with your code. Powered by LangGraph, NetworkX, and Ollama (qwen2.5-coder).

Evaluation module for the orq platfrom in python, including Red Teaming and Agent Simulation

Agentic QA boilerplate built on Playwright + KATA + TypeScript. Multi-agent skills following the agentskills.io spec, orchestrating the full QA lifecycle: shift-left planning, in-sprint testing, TMS documentation with ROI scoring, KATA automation, and regression with GO/NO-GO. Works across Claude Code, Opencode, etc.

Dialogue memory for AI agents (hermes plugin): auto-recall of facts across sessions, LLM fact extraction with a cheap gate, SUPERSEDE versioning. Local SQLite + hybrid search. One-command install.

Run your website by talking to it — the open-source, self-hosted AI-native CMS. Built on Drupal; export the output as a static site and deploy it anywhere.

A repo-native coding-agent loop harness for Claude, Codex, Copilot, and other CLI agents. Agents can edit. Gates decide what lands. You set the plan in motion. The agents take a prompt, and each iteration must update specs and commit through guardrails.

A curated list of benchmarks for evaluating Earth-Observation (EO) AI agents

Local-first AI project OS that turns coding models into reliable software agents with memory, sandboxed execution, verification, recovery, GitOps, and deployment.

증권사 애널리스트가 만든 한국 상장사 공시 분석 MCP 서버 — 재무제표·예상실적·주가·밸류에이션·지배구조·주주총회 의결권 분석을 Claude·ChatGPT 등 AI 에이전트에서 바로 무설치 사용.

A curated paper list on On-Policy Distillation (OPD) & On-Policy Self-Distillation (OPSD) for LLMs — student rollouts + teacher feedback, covering white-box/black-box methods, OPD-RL hybrids, agentic & multimodal applications, and industrial multi-teacher recipes.

43 Claude Code skills for making content that doesn't read like a machine made it — prose editing with an offline linter, prompt engineering for 40+ image/video/music models with optional one-command execution, and orchestrators from research to a published carousel, reel or proposal. Russian-first, English throughout.

Plasmod: An Agent-Native Database for Cognitive Objects, Event-Driven Materialization, and Structured Evidence Retrieval

Turn your Hermes agent into a financial analyst — on-demand equities / macro / options / prediction-market analysis, plus a standing watch: morning brief, earnings, 13F diffs, Polymarket. Powered by LLMQuant Data.

채용공고(JD)를 지원·면접·설명 가능한 IT 인턴 프로젝트로 바꾸는 한국형 AI 스킬 · JD → 오픈소스 선정 → 코드 분석 → 이력서·면접팩

PosterChan AI is a Nostr-powered personal cloud and a self-hosted AI powerhouse, running on one box you own.

Markdown 长文渲染成小红书 3:4 图片组的 Agent Skill，全文不删减，内置 5 种风格（Cursor / Claude Code / Codex 通用）

智简Copilot：面向大学生与青年求职者的 AI 求职工作台。简历编辑、JD 匹配、自荐信、投递看板、模拟面试与复盘报告一站完成。

A Perplexity API MCP server that unlocks Perplexity's search-augmented AI capabilities for LLM agents. Features robust error handling, secure input validation, and transparent reasoning with the showThinking parameter.

Run Geoffrey Huntley's Ralph Wiggum loop inside Pi, no shell scripting needed. Native commands control the loop, and the bundled agent skill writes the whole plan from a simple goal. Better than Codex goals and Claude workflows!

看不完的视频，让 AI 替你看懂：X/YouTube/B站视频 → 中文卡片解读（read-anything 的姊妹）

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

A multi-skill agent suite that gives any AI tool senior ServiceNow architect, governance, delivery, and development knowledge.

CAPO: Critic-Guided Action-Aligned Policy Optimization for Advancing LLM Agent Capabilities

Multiplayer admin and orchestrator platform for heterogeneous agent runtimes — Hermes (container-per-agent) and Letta (memory-first agents on a shared server) in one control plane, with multi-tenant security, group approval, and audit.

基于“躺着学”长期雅思教研与学员服务实践整理的 IELTS Agent Skills，覆盖学习计划、作文批改、口语陪练、阅读与听力复盘、词汇和模考复盘，支持 Codex、Claude Code、Cursor、WorkBuddy 等本地 Agent。

Open-source agentic software development platform for building, orchestrating and monitoring autonomous AI coding agents.

Markdown 长文渲染成小红书 3:4 图片组的 Agent Skill，全文不删减，内置 5 种风格（Cursor / Claude Code / Codex 通用）

An autonomous coding agent that owns its whole loop — engine, workbench, and memory. Apache-2.0.

Claude Code agents and skills for automated Playwright E2E test generation, debugging, and planning using the Page Object Model pattern

Local AI code review for coding agents. Rules enforced inside Claude Code, Codex, and Cursor. Free, open-source, fully local.

Chat-first trading agent on Agno + OpenAlgo: full REST coverage, 100+ technical indicators, and a human confirmation gate on every order.

Cursor for Reverse Engineering

An autonomous AI Research Agent Telegram bot powered by NVIDIA NIM (Mistral Large 3), DuckDuckGo search, and BeautifulSoup for real-time web scraping and comprehensive report generation.

⚡ ƝØVΛ — Free & open-source AI Agent for Telegram. Builds & hosts web apps, generates games, images, PDFs & voice. 100% free tier. Self-host in 10 min. AGPL-3.0.

Math Modeling AI Agent Flow | 数学建模全流程 AI 助手（国赛 CUMCM / 美赛 MCM-ICM）。四阶段流水线 Solver→Paper→Finalizer→Growth：解题认证、论文写作、提交终检、赛后复盘。防 AI 幻觉、防版本冲突，可安装到 Claude Code / Codex。

Rule-driven AI job search and application workflows for Codex. Your apply Buddy!

Portable Godot 4.7+ game-development skills for Codex, Claude Code, Cursor, and VS Code: GDScript, gameplay systems, shaders, optimization, publishing workflows, and AI studio orchestration.

DeepSeek Harness plugin: 15 obra/superpowers engineering skills, bilingual descriptions, per-skill toggles | DeepSeek Harness 插件：15 个 obra/superpowers 工程纪律技能，技能描述中英双语自由切换，每一个技能本身自由开关

LLM-driven session retrospection skill: detects friction in agent sessions and materializes learnings into correct destinations (user-memory, project-rules, skill PRs, checkpoints, harness artefacts)

Application framework for building AI agents with Pydantic AI. Used to build the next-generation agent system of YouWare.

Comprehensive toolkit for AI-assisted development: Claude Code permissions (400+ tools), Git workflows, and automation hooks

Train AI coding agents with RL (GRPO) on Prime Intellect — works with Pi Agent

A multimodal, agentic AI framework that captures and shares practical expertise by linking written instructions to real-world laboratory work. It uses video analysis to automate documentation and provide personalized guidance, aiming to making complex, hands-on science more accessible and reproducible.

A technical specification for the Jentic API AI-Readiness Framework (JAIRF). Defines the signals, dimensions, scoring model, and normalisation rules used to evaluate how interpretable, operable, and trustworthy APIs are for AI systems and autonomous agents.

Build and reuse WebMCP tools in a macOS desktop workspace for DeepSeek Harness.

The agent harness you steer — structured context, automatic staleness gates, and block-by-block review, all bound to your code. Local-first, MCP-native, open source.

Create an IRC chat bot powered by AI, using llamafile, in minutes.

Interactive Windows 11-inspired developer portfolio for Bharani Kumar S, built as a desktop-style web experience with verified projects, live data, and a grounded assistant.

Official repository for MiniAppBench. Contains the complete pipeline and codebase for LLM-powered interactive HTML generation and agentic evaluation (MiniAppEval).

🪐 Agent2World: Learning to Generate Symbolic World Models via Adaptive Multi-Agent Feedback

Local context engine for AI coding agents. Routes tasks to relevant files, tests, rules, and skills, supports prompt caching, and builds compact context packs for Claude Code, Codex, Cursor, MCP, and more.

Deep-dive notes and source code analysis of Claude Code and AI agent harnesses. Exploring memory mechanics and internal architectures.

A production GenAI architecture guide for Forward Deployed Engineer (FDE) and Applied AI interviews—RAG, agents, evals, security, and operations.

A shared AI agent for your team. Everyone trains it together; everyone talks to it privately. Self-hostable, FSL-1.1-ALv2.

Make destructive AI-agent actions reversible by default — quarantine + audit + human escalation for rm/git destructive operations. Reliability infrastructure, not a sandbox.

Extends Gephi with AI capabilities. Includes a Gephi plugin for graph analytics, an MCP server bridging AI assistants and Gephi, an agentic AI plugin with slash commands and workflow skills, and a specialized agent for network analysis research. Built for researchers working across network science and AI.

A semantic, transactional, replayable MCP control plane that lets autonomous AI agents operate Dwarf Fortress as a long-lived civilization. Safe-Rust phase-0B scaffold: MVCC world state, witnessed semantic plans, evidence-backed effects, deterministic replay.

在 Obsidian 侧边栏中使用本机 WorkBuddy：多任务、选区编辑、资料引用、来源追溯与本地文件上传

One command to keep secrets out of AI (LLMs). Works with Claude Code, Cursor, Copilot, Windsurf, and any AI coding tool.

ADOS Drone Agent — Software-defined drone network agent. Install on any companion computer to make your drone smart, autonomous, and fleet-ready.

MCP-based code knowledge graph and budgeted context subgraph retrieval for AI coding agents.

CADPilot is an MCP server that lets AI clients (ClaudeCode, OpenCode, DSH, ZCode) pilot FreeCAD, building constrained sketches, parametric features, and persistent-joint assemblies from natural language.

This GPT is Overpowered but Overly Cautious. / 这个GPT明明超强却过分慎重。

可扩展的插件化 QQ 机器人框架：多协议多账号、逐群功能开关、Web 管理后台与插件市场

LLM-powered MCP server for building financial deep-research agents, integrating web search, Crawl4AI scraping, and entity extraction into composable analysis flows.

Claude Skill for Generative Engine Optimization and agent readiness. Get cited by ChatGPT, Perplexity, and Google AI Overviews. Expose MCP, llms.txt, and OAuth to AI agents.

The official repo for BannerAgency: Advertising Banner Design with Multimodal LLM Agents

Multi-agent AI navigator that routes people in crisis to housing, food, healthcare and legal aid. 7 walkers, 40 real resources, any language. Built in Jac.

数学建模全流程引导智能体：三角色一体化（布鲁姆L1-L8引导式学习、苏格拉底深度探究、P-M-A-D四阶段建模引导），只做方法论引导、不代求解不代写

轻量、松弛、有效的 Coding Agent 配置套件。适合中等规模的 AI 原生项目开发。

Batteries-included, model-agnostic coding agent with native performance and a small runtime footprint

Long-term memory plugin for DeepSeek Harness (dsh): four-channel retrieval fusion and continuous memory evolution. Memory construction and retrieval are based on MemGAS (ICLR 2026).

AgentLabKit 是一个用于构建、运行、观测和评估 AI agents 的 Python 与 React 平台。它包含可复用的 Agent Runtime、provider-neutral LLM gateway、RAG engine、guardrails、workflows、long-term memory、evaluation、cost analysis，以及 Web/desktop clients。

Agent libOS provides a runtime substrate for self-evolving LLM agents with process-like execution, tool management, and security boundaries implemented through explicit capabilities and runtime primitives.

Code and Data for Paper "EvoArena: Tracking Memory Evolution for Robust LLM Agents in Dynamic Environments".

Build, share, and talk to autonomous Voice Agents powered by Vapi & ElevenLabs. (Built as a part of Vibe Coding Hackathon)

Build production-ready AI agents with the Strands Agents SDK and AWS services. This repository demonstrates how you can create multi-modal systems with persistent memory in minimal code. Progress from your first agent to production-ready systems through hands-on chapters.

MAIC · Multi-Agent Interview Coach

Animated macOS desktop pet and AI coding agent monitor for Claude Code, Codex, Cursor, and Pi, with local and remote SSH session tracking.

本地优先的 AI 求职工作台：采集与管理 BOSS 直聘岗位，基于真实经历匹配 JD、定制简历并准备面试，支持 MCP 接入 Claude Code、Codex 等 Agent。

A Benchmark and Agentic Framework for Omni-Modal Reasoning and Tool Use in Long Videos

Agentic framework combining the power of LLMs with domain-specific tools for materials science, enabling property extraction, simulations, and crystal structure generation to accelerate innovative material design and analysis.

Agent Skills. A collection of specialized expertise, workflows, and resources for LLM agents.

An agent runtime that treats context, execution, and permissions as OS primitives — not LLM problems.

Flowwink is the next gen open source Business Operating System with 500+ skills where Hermes agents, Claude, OpenClaw or internal FlowPilot run CMS, CRM and ERP autonomously! Help establishing FlowWink as the self hosted, private LLM, open source agentic BOS alternative!

[EMNLP 2025 Oral] IPIGuard: A Novel Tool Dependency Graph-Based Defense Against Indirect Prompt Injection in LLM Agents

A general-purpose robotic agent framework based on LLMs. The LLM can independently reason, plan, and execute actions to operate diverse robot types across various scenarios to complete unpredictable, complex tasks.

Open-source protocol integrating collective and artificial intelligence — LLM orchestration, agentic workflows, and RAG in a cross-platform TypeScript/Tauri app.

Official repository for “From Trainee to Trainer: LLM-Designed Training Environment for RL with Multi-Agent Reasoning”“

AI-native multi-agent marketing platform developed with FastAPI, Google ADK, and React JS. Orchestrate 70+ specialized agents powered by Gemini 3.1 and Veo 3 for automated branding, SEO, and high-fidelity content creation.

🤖 A curated collection of awesome resources, tools, skills, and projects for nanobot - the ultra-lightweight AI assistant framework.

AI Agent eBPF optimization benchmark and framework

A collection of Agent Skills for building iOS/macOS applications with Apple's Foundation Models framework (Apple Intelligence). These skills are designed for zero-friction AI-assisted development with on-device models.

LLM-first SEO skill for AI coding assistants. Features 16 sub-skills, 10 specialist agents, 89 python scripts, and 160+ markdown marketing skills.

A self-hosted web interface for pi coding agent.

Single-file AI coding assistant (~350 LOC). Claude API with tool calling. TypeScript + Bun. Zero dependencies.