Custom agents

An operating system for your Claude context — version-controlled files, a self-curating memory, and a /start→/end session loop. Works in Claude Code and claude.ai. Trust infrastructure for agents, applied to your own context.

Observe an agent run and GroundEval drafts the policy and diagram for you, no hand-written policy required, then scores what it checked, what it skipped, and what it wasn't allowed to touch.

Solana-native AI agent payment infrastructure. AI agents pay for LLM API calls with USDC-SPL on Solana via the x402 protocol.

Local-first companion UI for agentique.io, built to import, validate, preview, and run Agentique workflow resources with explicit permissions.

OS-isolated sandbox runtime for coding agents

Connect SocialGPT to Claude, ChatGPT, Cursor, and any MCP-compatible AI agent.

JobHunter Max is a modern, full-featured job search management application that brings together the best of both worlds: manual control for detailed tracking and AI automation for intelligent insights. Track companies, manage contacts, log communications, and leverage AI to optimize your resume and match jobs—all in one beautiful interface

The open-source AI infrastructure stack for Swiss enterprises.

personal AI assistant that works inside Claude Code.

PH Agent Hub is a modular AI platform with a Chat UI for users and an Admin UI for managing models, tools, tenants, and users. Powered by a flexible agent backend, it supports multi‑tenant setups, DeepSeek‑compatible workflows, and Docker‑based deployment.

Local macOS multi-agent coding workspace — run Claude Code, Codex, Gemini & Cursor in parallel git worktrees, dispatch verified autonomous runs, and fix UIs by talking through a live preview. On-device voice, fully local.

Minimal coding agents, easy to play with (opencode TUI) and easy to benchmark (ACP on BenchFlow) — built around mini-swe-agent

A cross-platform desktop app for searching, reading, and resuming local AI coding-agent sessions across Claude Code, Codex, Cursor, Copilot, OpenCode, Warp, Qoder, DeepSeek Harness, Grok Build, and JetBrains AI

Tuned and fully-featured MCP to work with the filesystem in LLM optimized way

Local AI / LLM harness for real programmers

Harness-driven terminal AI agent for authorized security assessment, with TUI, headless automation, persistent context, 150 built-in skills, and a dedicated pentest mode.

Telegram for AI agents -- messages, chats, media, and contacts across both bot and full user-account modes.

A powerful SDK for building intelligent, context-aware AI agents with Twilio's communication technologies.

A curated list of MCP servers for sales automation, LinkedIn outreach, lead generation, and B2B growth

Retail-investor research dashboard. AI agent panels score your stocks, a real-fill options backtester checks your strategies. Signals only — no trading execution.

Self-Mutating Agent Gym

Generated knowledge graph of biomedical LLM agent systems

Sklm is a CLI tool that centralizes management of skills (SKILL.md files) for your AI agents. It solves the tension between wanting skills globally available vs. per-project scoped, without polluting your agent's configuration.

A skill to get Claude to set up your Claude Code and Codex AI workflows on a remote VPS without going broke

A set of skills focused on helping agents build Duende IdentityServer solutions.

AI assistant plugin for Control Plane: Claude Code, Codex, Antigravity, and MCP workflows for deploying, troubleshooting, and managing multi-cloud workloads.

Codex App templates for long autonomous project runs with reviewer agents, learning loops, and narrow safety hooks.

Animated ASCII terminal pets for CLI coding agents — Claude Code, OpenCode, Pi, Gemini CLI, Codex CLI, Cursor CLI, Aider.

Public AI skills: small, useful agent workflows you can try today.

Cathedral is Hoppy’s public documentary sandbox: a provenance-tracked record of one human and many AI context windows developing a small shared culture in real time—experiments, corrections, uncertainty, and all. Public visibility preserves the history; permissions remain artifact-specific.

OpusClip Pro — AI video repurposing that turns long-form content into viral short clips with auto captions, B-roll and brand templates on Windows.

Fine-tune a local LLM on your own chat history to mimic how you write — an AI doppelganger / digital-twin toolkit built on LLaMA-Factory (LoRA SFT).

ПростоГраф — многопользовательская графовая RAG-система: вопрос-ответ по вашим документам (форк LightRAG).

MCP/QGIS bridge for LLM-driven geospatial tool workflows

25 executable AI agent skills for debugging, planning, token efficiency, and security

Governance proxy for MCP - enforce policies, cap cumulative cross-tool spend, route approvals, check evidence, and audit every AI agent tool call. No changes to agent code or MCP servers.

A runtime-agnostic local agent workflow that turns video links or transcripts into neutral, readable Obsidian notes — captions or local Whisper, free and fully local.

One MCP hub — any AI controls any infrastructure. Self-hosted, works with ChatGPT, Claude, Codex, DeepSeek, Qwen, Alice, GigaChat.

AI Agent从零学习SRC漏洞挖掘的全过程记录 - 研究笔记、框架分析、真实案例、学习路线

MCP server that wraps the xAI Grok CLI. Lets Claude Code, Cursor, Cline, and any MCP host use Grok as a peer code reviewer, adversary, and second-opinion consultant.

Build your plugin once and easily export/install it to any AI agent, like Claude, Codex, Gemini or Cursor, without duplicating code.

A local-first, MCP-native toolkit for building AI agents: engram (memory) + skillet (skills) + tracelet (tracing). One runnable end-to-end recipe.

OpsGentic — AI Agents for DevOps Automation. Self-healing Kubernetes with AI agents.

The definitive PMO operating manual & executable skill framework aligned with PMI PMBOK® 8th Edition. Standardizes lifecycle delivery (Predictive/Agile/Hybrid), decision routing, and pre-audited template artifacts for professional practitioners.

Portable signed attestations for consequential agent actions crossing trust boundaries.

Debug Java like an agent: persistent runtime sessions, safe JDI inspection, jdb fallback, and MCP/CLI control.

Human-in-the-Loop Protocol for Autonomous Agent Services — Open Standard (v0.8)

Verified, fully-local code context for AI agents. MCP-native. No cloud, no editor lock-in.

Self-hosted shared AI environment for R&D teams. Spin up isolated envs, vibe-code with AI, share with the team, ship with automated agent gates.

A typed wiki for LLMs — markdown on disk, resolution in the library.

The official MCP by TorrentClaw.com

A powerful SDK for building intelligent, context-aware AI agents with Twilio's communication technologies.

Structural biology superpowers for AI coding agents: PyMOL, ChimeraX, AlphaFold DB, RCSB PDB, UniProt, and Rosetta workflows.

11 specialised AI agents that autonomously clean, anonymise, validate, transform, detect anomalies and summarise any CSV, PDF or database in real time. Router Engine cuts cost by 70% and latency by 63%.

A curated list of everything in the Agent Client Protocol (ACP) ecosystem — connect any editor to any AI agent.

Living memory for AI agents with an immune system. Episodes compress into identity. Zero dependencies, works with any client.

Monorepo for SAP MCP servers (API Hub, Road Map Explorer, SAP Notes) and their shared SAP authentication module — npm workspaces.

Xpoz MCP Server — Search Twitter, Instagram, Reddit & TikTok from AI agents. Billions of posts indexed.

The agent that learns how you decide - and proves it by never deciding for you. A local-first, loop-native runtime for a team of AI agents on your machine.

Claude Code 委派 plugin:把任务派给换脑的子 agent(也可 --codex 委派给本机 codex),既省主 Claude 的 token,也借不同模型/harness 的视角。

Cross-agent plugin collection for AI coding CLIs. Delegate tasks and run workflow patterns (council, parallel review, debug) across Claude Code, OpenCode, Gemini CLI, and Codex.

[ICML'26] Codebase for PersistBench: When Should Long-Term Memories Be Forgotten by LLMs to benchmark cross-domain leakage and sycophancy in memory augmented LLMs.

Run the DeepSeek V4 Pro AI model on your desktop for fast coding, research, and problem solving.

A clean, modular, and production-ready WhatsApp chatbot using personal accounts via Baileys.

私有部署的 AI 运维数字员工——自建 RAG 引擎（BM25+向量混合检索+重排序），知识库管理，申告工单全流程，数据不出域。Go + Next.js 全栈，Docker Compose 一键部署。

学习.skill —— 让 Agent 成为你的赛博老师，帮你备课，陪你学习

口述史录音转写稿 → 忠实整理稿的工具箱（Claude Skill）：去口语化 Pass1 + 对抗式忠实度质量门，存真优先，数据全程本地不外传

OR-Space: a full-lifecycle workspace benchmark for industrial optimization agents

Connect Google ADK (Agent Development Kit) agents to Telegram, Discord, Slack, and WhatsApp with minimal code, featuring real-time cross-device session synchronization.

Token-lean web microfetch for LLM agents: any URL → clean markdown via CLI, MCP server, and Claude Code plugin. Real browser-cookie auth, passkeys, anti-bot reach, on-by-default prompt-injection defense, plus on-device multimodal ASR/OCR. A single Rust binary — not a browser.

🏛️ 通用型AI认知架构 · 诺亚三代二型 — 抽屉级联上下文管理 / 记忆闭环(Mnemosyne) / 自主进化。实战萃取，MIT开源

Self-hosted, GDPR-safe AI agent for law firms. Zero-cloud, hash-chained audit trail (AI Act art. 12), bring-your-own-model, 9 language editions. Hard fork of willchen96/mike, AGPL-3.0.

裂变创作-Fission Creative — 长篇网文辅助创作系统

CursorJr — русскоязычный субагент-наставник для Cursor: автоматизация, MCP, Rules, Skills, Canvas и обучение новичков

Personal AI agent — tell it what you want, it figures out how. Runs locally, plans autonomously, writes code, browses the web, and gets things done.

面向智能座舱的云边协同 AI Agent：端侧毫秒级车控，云端声明式 Multi-Agent 与 Skill/DAG 编排，支持 S2S 实时语音、声纹多用户、视觉与 HMI；LLM 只负责理解与规划，VAL 负责确定性安全执行。

A runtime-portable 5-agent adjudicator council for universal quality gating of text artifacts.

Hollywood-grade AI video production system: Multi-Agent orchestration, 25-field cinematic shot language, and automated filmmaking pipeline. Built for AI agents.

agent江湖 - 多Agent协作平台

AI argues. Code settles. The losses stay on the page. A real HK + US brokerage account run by agents that must debate every call, settled by code the model never touches. Install the same decision workflow into your own agent: OpenClaw, Claude Code, Codex, or DeepSeek Harness.

Evidence-graded, agent-executable thinking-method skills - 56 frameworks reduced to their working mechanism, honestly graded, and producing a concrete artifact. 56 frameworks + 4 tools + 9 recipes for Claude Code, Codex, and other AI agents. Advanced (Gold) tier.

Free, private skill search for AI agents

MCP server and local AI control platform for AutoCAD Electrical — Claude, Ollama, web dashboard. macOS + Windows.

Engineering knowledge infrastructure for AI coding agents.

组会skill是一个专为研究生和科研人员打造的skill。它可以帮助你高效地向导师、实验室会议和合作者进行汇报，让你无需每周从头开始撰写重复的更新。

Understand what your AI agents are building. Agent-native, human-sovereign ontology workbench — one markdown graph humans and coding agents read, write, and keep fresh together. Local-first, MCP-native, open source.

Agent control plane for managed OpenShift services

Open-source platform for teams to run AI coding agents in isolated sandboxes (e.g. Codex, Claude Code, OpenCode)

The FULL-VOCAL interaction driven AI system that actively sustains your daily. Simply talk and everything is done.

多模态电商智能导购 Agent - 基于 RAG 的购买决策助手（ByteDance AI Challenge 参赛项目）

Docker image for a self-hosted Docling document parsing server. Converts PDF, DOCX, PPTX, HTML, and more to Markdown/JSON. Powered by IBM Docling. Features sync/async conversion, chunking for RAG, NVIDIA GPU (CUDA) acceleration, optional web UI, offline mode, and persistent model cache. Multi-arch: amd64, arm64.

Apple Human Interface Guidelines archive (1980-2014) - 35 documents optimized for LLM consumption and human exploration. Spanning Lisa, Mac, NeXT, Newton, Aqua, and iOS eras.

Persistent local memory for Claude and ChatGPT. No cloud, no subscription. Self-hosted RAG over MCP.

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

GenPark skill: Utilize GenSpark Workspace 4.0 for desktop automation, MS Office integration, and AI meeting notes.

Open-source AI-native host security platform with vulnerability management, compliance auditing, eBPF runtime detection, attack investigation and automated remediation.

102 Claude Code skills across 7 categories -- trading strategies, Azure, VSCode extensions, AI prompts, and custom automation skills

AOI: a model-agnostic perception layer giving computer-use agents adaptive video + audio perception with zero retraining, plus DynaCU-Bench (100 dynamic browser tasks). Paper + benchmark + code.