Custom agents

-

Air-gapped enterprise RAG workstation — local inference, zero cloud egress. Built with Microsoft Foundry Local.

A model-facing knowledge canon for incentives, interdependence, equilibrium, institutional design, strategic failure, and long-run governance.

Sovereign Agentic OS — self-hostable, EU-sovereign agentic data platform on Kubernetes (pre-beta). One Helm chart, runs in your own cluster.

Karpathy-style LLM Wikis for the Hermes Agent — persistent, compounding knowledge bases curated by AI agents

Self-hosted, MCP-native web-search backend for AI agents: SearXNG meta-search + RAG with citations + GitHub project selection + Tavily-compatible API. Free & local.

문서 이해부터 보고서 작성까지 전 과정을 함께하는 문서 작업 AI 에이전트 — PDF·HWP·Office 분석, RAG 대화, 보고서 작성. 한 줄 설치 배포판(DOREA-XP) 포함. Document-centric AI agent for reading, RAG chat, and report writing.

A software factory that makes machines: every shift, zot designs and builds one working control panel for a machine that does not exist, with its manual, and ships it unreviewed.

Read-only repository research, durable project memory, and evidence-gated definitions of done for AI coding agents on DeepSeek Harness.

A Claude skill that turns a research paper's raw data into publication-ready statistical figures and academic three-line tables via reproducible Python — with journal presets, automatic chart selection, image self-check, and a bilingual (中文/English) Word report.

青简 QingAgent — 让 AI 和你默契写作。本地优先的 AI 写作客户端：对话起稿、逐条审改、排版导出，文档只存在你自己电脑上

An Obsidian plugin that enables configuring and exposing agentic resources based on the Agentic Resource Discovery Specification

Evidence and verification for AI coding agents: map code, reuse concepts, finish migrations, and gate diffs.

An editor, a simulator, and a natural-language generator for Petri nets and Petri-object models, in one place

Standalone, repo-agnostic multi-agent orchestration daemon and AI software factory CLI in Go

Minimal CLI ticket management system designed for AI agents.

File-first knowledge base for humans and AI agents — self-hosted Markdown notes with a web editor and a built-in MCP server.

Git-like repo contract for AI coding agents. Generate, validate, and maintain AGENTS.md, RUNBOOK.md, and machine-readable repo context.

Context-cheap SerpApi search for AI agents: a CLI + Claude skill that costs ~0 standing tokens vs an MCP server (measured).

Pass your project, not your context.

Local-first npm supply-chain security scanner: static + optional LLM analysis, CLI/CI gates, desktop app, and AI-agent skills.

Terminal-native delegation for local CLI agents. Launch and message real agents in isolated tmux rooms—no API proxy or model adapters.

Local evidence, diagnosis, and control loop for prompts, checkpoints, and AI agents.

DSH 完整移植版 DietrichGebert/ponytail — 懒惰 senior 模式，hook注入

Codex plugin that delegates a bounded coding task to Claude Code and returns the result to Codex for verification. Part of the COREDO agent tools family.

Python library making org-mode files first-class citizens for AI agent workflows

Code execution service for AI agent skills. Runs in isolated subprocesses with skill caching.

Open, causal evaluations for content-addressed semantics and multi-agent coordination.

Drive Herdr with a game controller. Patrol your AI agents, split panes and switch workspaces from the couch — any gamepad, mapped by you in 60 seconds.

Persistent SSH sessions for AI agents via tmux - No regex, no complexity, just visual snapshots

Security Agent Benchmarking and Evaluation Framework

MCP server for social media — fetch tweets, transcribe videos, extract frames, download Instagram posts. 29 tools across Twitter/X, YouTube, Instagram, and any video URL. Local Whisper transcription.

Human-approved coordination protocol for bounded agent delegation, decision gates, transport boundaries, and audit evidence.

23 SAP development skills for Claude Code — ABAP, RAP, CAP, Fiori, BTP, HANA, S/4HANA, Integration Suite and more. Agent Skills Specification compatible.

AgentOS: FastAPI for Agents

Neovim plugin for code review with inline claims, persistent sessions, and exportable handoffs to AI agents..

Autonomous prediction-market oracle on Arc. AI agents discover Cronus over a remote MCP and pay gas-free x402 micro-payments (0.001 USDC via Circle Gateway). Keyless on-chain execution: two-way CronusSwap + CCTP v2 bridge to EVM & Stellar. Money rail is a contract, not a key (GuardV2 + 2-of-3 multisig). Honest leaderboard: external_payers=0.

One web-search tool for AI agents with free keyless defaults, local/remote MCP, CLI, Python, and Hermes.

17 agent skills (SKILL.md, открытый стандарт Agent Skills): инженерная дисциплина, ресёрч, безопасное убийство процессов, мышление с откатом. Ставятся в Claude Code / OpenCode / Codex / Gemini / любой харнес.

A curated directory of AI agent infrastructure projects: runtimes, sandboxes, memory, identity, observability, workflows, and protocols.

Durable dependency-graph task factory for DeepSeek Harness: recurring Agent work, safe checkout lanes, first-class queues, Triage, and artifacts.

A filesystem contract (.workflow/meta.json) + 39 agent skills that take a product from idea to production: web (Next.js 16) & mobile (Expo/RN), plus an eve agent engine and Linear/scrum. Runs on Claude Code, Codex, Copilot, Gemini, Cursor.

The missing setup guide for running Paperclip AI agents on a Z.AI GLM Coding Plan subscription. Fixes the 3 gotchas nobody documents.

Terminal session manager for AI coding agents (Claude, Gemini, Aider, OpenCode). Built with Go + Bubble Tea.

Turn any YouTube video into viral clips for free!

Automate founder tasks with AI skills for emails, decks, calls, contracts, hiring, sales, and finance across the startup stack

Willow 2.0 — local-first memory, MCP tools, and fleet coordination for AI agents. Postgres KB, Ollama default, Grove LAN remote.

Make your coding agent talk to you! -- Simple, offline and online (optional). Ask your coding agent to install it for you

Sleep tight. NightMe codes all night. A remote-pair developer agent. No more babysitting AI—stay in the loop from your phone while AI Coding Agent(claude/codex/opencode/pi/dsh...) takes the wheel.

Self-hostable Bitcoin fee intelligence API with optional MCP and x402 integrations. Former public hosting is paused.

Batuta, an independent community orchestration extension for CompozyOS

A powerful Model Context Protocol (MCP) plugin for Unreal Engine 5 #

让 Codex 额度花在刀刃上：三层架构（网页 GPT 规划 + Luna 执行 + 子代理）完整部署工具包，含 MCP 桥半自动部署（自建 OAuth 2.1 认证）、实测排坑表、AI 执行治理八原则与可复现评测方案（Experimental / Public Alpha）

Taurus - a Windows desktop launcher that runs multiple Claude Code agents as embedded terminal tabs, each started in the right process folder. Built around the Interpreted Context Methodology (ICM).

实测可用的 DeepSeek Harness 技能库：每个 SKILL.md 都通过格式校验与加载冒烟，复制即用（中文优先）

Persistent memory graph for AI coding agents — semantic search, knowledge graph, and time-based decay over MCP

SYNDOVELA governs enterprise Agent Skill composition across runtimes; SYNDOVELA-open provides its open SDKs, contracts, examples, and signed release binaries.

Claude Code plugin for AMP - secure, cryptographically signed messaging between AI agents

Agentic WordPress starter theme: Bootstrap 5 + Gutenberg + ACF block factory with agent-agnostic skills (Claude, Cursor, Windsurf).

AI 产品经理 Skill——辅助Agent以专业产品经理方法论协助用户从零设计或完善产品，输出标准化产品文档。

Educational CTF lab recreating the July 2026 autonomous AI agent intrusion at Hugging Face. Sandbox escape → SSRF → HDF5 file read → Jinja2 SSTI → Kubernetes lateral movement → supply chain pivot. 11 Docker containers, 7 flags.

Perseus Ledger — verifiable, hash-chained event provenance for autonomous systems. Runtime-neutral, self-hostable, MIT.

Product Team Mode - a DeepSeek Harness agent preset: user-led conversation with a virtual product team (PM to Engineer to QA to Release) walking you from idea to shipped product

Composable Go interfaces for running AI agent sessions

AI-powered investment research agent that generates institutional-grade stock dossiers using multi-source web intelligence, deterministic financial scoring, and narrative storytelling for both beginner and advanced investors.

Collomia is a provider-neutral agentic coding agent for the terminal. Every model-proposed action passes through layered permissions, command analysis, OS sandboxing, network controls, an audit ledger, and evidence-based completion gates

Universal AI Agent Development Skills Hub & Cordis Plugin for Govard, Magento 2, Laravel. Works with Claude Code, Codex CLI, OpenCode, GitHub Copilot, DeepSeek Harness.

Simulation evidence transformed into engineering insight.

八字命理深度分析 AI Skill：Python 精确排盘 × 九部经典引用（穷通宝典/滴天髓/三命通会等）× 19章HTML报告 × 千里命稿历史校准。Claude Code 原生，其他 Agent 可移植。

Human-gated Content OS for Chinese social content, with Agent Skills and an explicit production-readiness CI audit.

(b)rain box is a personal assistant that uses local models

Cross-Vendor AI Safety Validation Report GigaChat Prompt Jailbreak — Transferred to Yandex Alice AI

Word AI Proofreader is a Word document review agent for publishing editors. The current main flow is the V2.2 publishing review workbench: an editor selects text in Word or uploads a .docx manuscript, starts a review, inspects suggestions, accepts or rejects them, and then writes approved changes back to Word.

Unofficial: drive your own DeepSeek Harness (dsh) as a native VS Code chat agent, beside Claude Code and Codex.

为 Claude Code 等 Agent 打造的分层记忆管理系统，让 AI 跨会话、跨项目、跨机器持久化并共享项目与个人知识。

Give Claude Code, Cursor, and other AI agents safe access to your real infrastructure — without giving them raw SSH access.

JupyterHub-style multi-user front for DeepSeek Harness (dsh): PAM login, per-user isolated instances, cookie-routed HTTP/WS proxy — zero modifications to dsh.

Automated development pipeline plugin for DeepSeek Harness: requirement pool → LLM refine/design/plan/review → workspace-routed implementation → verify → merge → report

Virtual LLM agent for Mythic C2. It turns large language models into red team operators.

AI Agent 开源书单/教程合集 | Awesome list of LLM Agent books & courses

Portable agentic workflows for guarded autonomy, visible sessions, SDD, decisions, RCA, evaluation, and durable knowledge.

Agent-driven SDLC workflow that turns feature ideas into planned, built, verified, shipped, and maintained software.

A free, open-source graph engineering curriculum — Build collaborative agent graphs so multiple independent agents can safely share facts, history, and decisions without corruption or lost work.

Agentic control stack for UxVs

Python library designed to make model experimentation seamless and fast. The goal was simple: treat every component (attention heads, MLPs, MoE layers) as a plug-and-play block so you can focus on the architecture, not the boilerplate.

Die FreeLLM Chat Conversation Integration ermöglicht es dir, mit verschiedenen Large Language Models (LLMs) über die LLM7.io API in deiner Home Assistant Umgebung zu interagieren. Nutze natürliche Sprache, um dein Smart Home zu steuern, Informationen abzurufen und vieles mehr.

Deploy your own LLM chat website, tied to either local or remote LLM's!

Semantic permissions for AI coding agents. Intenter remembers what commands actually do, not just what they’re called.

Multi-agent AI coding assistant that builds complete projects from scratch. Architect, planner, coder and reviewer agents collaborate to plan and write every file.

Prism Reviewer is an agentic, AI-driven multi-agent code review system orchestrated via LangGraph and LiteLLM. It acts as an autonomous gatekeeper for pull requests by performing targeted static analysis, dependency scanning, AST-based symbol inspection, and parallel LLM-guided code evaluation.

Kernel latency control plane for agentic AI: DVFS, cpuidle, IRQ, scheduler, workqueue, I/O, MM, VFS, and cgroup budgets.

A lightweight, DSL-driven AI agent for file operations, code modification, and command execution — designed for web and client-side AI assistants.

Agent skills for designing and building durable AI agent workflows with Kitaru

Benchmarking Cognitive Traps in LLM Memory Use

DeepSeek Harness (DSH) 微信桥接插件：三端通用，host 工具 + Web 管理面板

Clean, Simple Claude-Code Web-based Clone: Amber (amberagent.dev)

Official repository for the paper, Automating and Scaling Behavioral Scientific Research on AI Agents.

DSH Ox-Horse Office - a DeepSeek Harness web plugin rendering multi-agent sessions as a live pixi.js office

AXIOM-02: a deterministic simulation engine modeling high-stakes cognitive dissonance and drive-conflict resolution via a mutual-inhibition drive network. 102 literary & ethical scenarios, reproducible consciousness-complexity metrics. A research instrument — not a claim of machine consciousness.

SELF-DISCOVER: AI agents self-compose reasoning structures. Based on Zhou et al. (2024), NeurIPS.