Custom agents

A collaboration space where humans and AI agents carry projects forward through shared context, durable roles, work, documents, meetings, and handoffs.

One-stop agent skill set for mathematical modeling contests: problem analysis, model building, algorithm implementation, publication-grade figures, and paper writing with scoring

Gemini, Grok and Codex as read-only research & review units inside Claude Code — on the subscriptions you already pay for.

Search and analyze global news coverage and US television transcripts via the GDELT Project's real-time APIs via MCP. STDIO or Streamable HTTP.

DeepSeek Harness plugin for 得到大脑（Get笔记）

Adaptive running and strength coaching for the long run — a hosted MCP server and Agent Skill that keeps one current 28-day plan in step with your Intervals.icu evidence. Independent project, not affiliated with Garmin or Intervals.icu.

Agent Skill，AI 驱动的学术论文工程化技能，核心功能为【Latex 排版】和【翻译】。可以将 DOCX、PDF、Markdown、LaTeX 等多种文档格式转换为可编译的 LaTeX 项目，内置 8+1 种期刊模板，也支持用户自定义Latex格式上传，还有学术汉译英以及 12 项问答功能。AI-powered academic paper engineering Skill: transform DOCX/PDF/Markdown/LaTeX into compilable LaTeX projects with 8+1 journal templates, Chinese-to-English translation, and 12-item QA.

Hermes Agent (Android) — 通过 proot + Ubuntu rootfs 在 Android 上运行 Hermes Agent，内置 WebUI 仪表盘。无需 root。

BrainAgent — brain-inspired plugin for DeepSeek Harness (dsh): a pipeline of heuristic filters, statistical memory stores and context injections; reward-ledger + UCB1 bandit learning loop (RL-lite). 915 tests. Free noncommercial use.

OpenSquilla Android: Chaquopy + FastAPI gateway packaged as one APK with embedded Vue WebView UI

An open-source terminal coding agent built on a bounded, observable Rust runtime.

RoughCut Agent — Autonomous video editing agent that runs intent → analysis → triage → blueprint → compile → review in one shot

AI-based phishing detector using 200K+ dataset, smart features, and ML model.

桌面端 Claude Code 工作台，支持会话级上下级协作（主管派活 / 员工执行 / 汇报验收）。

Pin, blitz, merge. Autonomous coding pipeline: GitHub issue → EC2 spot instance → OpenCode agent → pull request. From issue to PR. With a trail.

An orchestration DSL and embedded runtime for hosting agentic systems. Durable, inspectable workflows with deterministic step ordering, running in-process under a capability jail — the model lives in your host, not in the runtime. 30+ companion packages for memory, transport and governance.

Neovim-style TUI for DeepSeek Harness (dsh): nvim renders the terminal, a Node runner bridges DSH agent events over msgpack-RPC — sessions, tool cards, reasoning panel, approvals, markdown tables

openEHR assistant AI plugin suite with Skills, commands, and MCP server configuration for clinical workflow integration

The 100% Free & Open-Source Wispr Flow + Otter.ai Alternative. Local Whisper Voice Dictation, Dynamic Island & Notch UI, Screen AI Assistant & Notes for macOS & Windows.

Claude Code & OpenClaw plugin, with a context-aware rules engine that injects only the rules each prompt needs. Codebase-aware, ~4ms lookup, pure Python. 222 rules / 30 domains for full-stack and scientific work. Includes a deterministic security audit, automatic session handoffs, ranked project memory, context usage alerts, and review agents.

Open-source, self-hosted MCP server for project intelligence, semantic search, context engineering, and persistent agent memory for AI coding agents.

The memory is yours. The model is a guest. Local-first memory and retrieval layer for AI agents: RAG with no model in the path, abstention as a verdict, MCP server, evaluation included. Rust, Apache 2.0.

AI-native NoSQL + vector database with built-in semantic search for RAG & AI agents — the Dart & Flutter client. Self-hosted Pinecone alternative.

🍪 High-density codebase genome compressor (≤600 tokens) & Bayesian Darwinian memory evolution for autonomous AI agents.

MCP server over the Remnawave panel and SHM billing: 34 read tools and 15 mutators behind a plan → confirm → apply step. Either backend alone is a complete setup.

AI-native worldbuilding canon: turn natural-language conversations into structured, validated Markdown/YAML and interactive graph views.

用AI“炼化”你的数字人格（Persona）。导入微信/QQ聊天记录，一键生成复刻你语气与思维的人格Skill。万物皆可Skill。

270 business concepts from The Personal MBA by Josh Kaufman, structured as an AI agent skill library. Give your agent a complete business education.

Local-first semantic search across tens of thousands of agent skills (SKILL.md) for Claude Code, Codex & OpenClaw — describe your task, get ranked matches with install commands. FAISS + Ollama, no API calls.

AI-assisted network firewall, filtering resolver and traffic analyser. Little Snitch's per-connection visibility for the whole network, with UniFi's operational surface — plus YouTube ad removal, WireGuard, Tailscale exit nodes, and an assistant that can actually change the rules.

An infinite canvas for thinking with AI. Every branch is its own conversation with its own context and Models.

Context Management Layer for Agentic AI. Graph-based knowledge context, SQLite index, advanced FTS5 lexical/BM25 search and cross-session live memory.

This project is a compounding knowledge system ("Second Brain") that manages multiple Obsidian vaults, combining Johnny Decimal for structural organization with Zettelkasten for emergent meaning, augmented by AI librarian agents for maintenance.

CompAgent: An Agentic Framework for Visual Compliance Verification

Private, offline-first local AI agent & developer assistant. Runs embedded llama.cpp (GGUF) entirely on-device with WhatsApp, Discord, and terminal CLI tools.

Desktop app that runs agent CLIs in project-scoped rooms and presents their history in one shared view.

Agentic LLM shell aimed at low cost GPUs

S.A.K. Utility is a portable Windows toolkit for PC technicians and sysadmins, combining profile backup/restore, diagnostics, app migration, cleanup, network tools, ISO/USB imaging, email archive tools, and AI-assisted workflows in one repair-ready package.

OpenFOAM CFD automation Claude Code Skill set - simulate (of-sim), post-process (of-post), diagnose failures (of-doctor), one-command machine setup (of-setup)

ENGRAM — an epistemic memory substrate for AI agents. Public release snapshots (one squashed commit per version).

Local, read-only AI agent that investigates SIEM alerts (Wazuh) with an on-device LLM via Ollama and writes an enriched triage report for a human analyst

MCP server for Semantic Scholar API with 14 tools: paper search (relevance + bulk), authors, citations, recommendations & datasets. Supports stdio and streamable-http/Docker.

PyQT6开发的PDF论文阅读/翻译/AI总结小工具(仿知云) Vibe Coding作品

A check that cannot fail looks exactly like a check that passes. AQK plants a known defect into a copy of your most-repaired file, runs your own declared checks against it, and counts a catch only when one names the planted file. Zero dependencies, one command.

An offline AI chat application built specifically for Ubuntu Touch, allowing users to run lightweight Large Language Models (LLMs) directly on their mobile device.

Go client for the Twitch Helix API and EventSub with typed endpoints, examples, and CI workflows

Find teammates, build real projects, earn a verified reputation. Semantic matching recommends projects and people by skill, a RAG-grounded assistant answers from live platform data, and every finished project leaves ratings, endorsements and badges behind. Flask + PostgreSQL + Socket.IO — live demo included.

Toolkit agentic no oficial de JW: CLI, MCP, RAG, Skills, etc.

My Chisa wifey👅

Self-hosted platform that runs AI agents like employees: identity, Docker sandbox, brokered credentials and a backlog per agent, with guard rails outside the model. Go + Postgres, AGPL-3.0.

Local toolchain for evidence-linked knowledge graphs: normalize sources, validate, merge, evaluate and render graphs authored by an external agent or compiler.

An interactive, dynamic, and privacy-first AI developer roadmap for 2026. From zero to building autonomous agents, MCP servers, and micro-SaaS. Built with Next.js, Tailwind v4, PydanticAI, and Vercel AI SDK.

BelieveMe — evidence-first execution harness for safe, verifiable AI code changes

Desktop-ready DeepSeek Harness Ultimate: a curated, deduplicated plugin distribution that solves choice overload across coding, workflows, research, safety, and automation.

Universal digital fingerprinting MCP server — 103 techniques in 13 tools for TCP/TLS/SSH/HTTP/DNS/WAF/IoT fingerprinting, C2 detection, OSINT enrichment

AI agent skill for preparing Chrome Web Store submission materials

Chimera Proxy: cross-platform Rust/Tauri desktop client for VLESS, REALITY, XHTTP, TUN, DNS and Mihomo-compatible proxy cores.

Deterministic compiler for coding-agent config

Every Claude account in one menu bar — swap before you stall. Native macOS app for the claude-swap engine.

An open-source AI VP of Marketing for in-house B2B teams. Modular context portfolio, parent-orchestrator skills, two-layer memory, defensive gates. Companion to Stack & Scale Issue 42.

Unofficial local-first Withings MCP server for AI health, sleep and activity agents.

Knowledge that can show its work: an open, federated knowledge system for humans and agents.

The content studio MCP for coding agents — BYO-key generation, deterministic video/carousel rendering, draft-first social publishing.

Conversation Landmarks (对话地标) for DeepSeek Harness — fixed navigation and hover previews for long agent conversations.

Turn any link or Markdown into a publish-ready Xiaohongshu (XHS) asset pack: structured summary, 8-style SVG knowledge cards, borderless 1242×1660 PNG exports, plus post copy (title/tags/body/CTA).

The simplest CLI for shareable links to your web files. Password optional.

Local-first log dashboard and error tracker for Laravel, Horizon, nginx, and console logs — the Laravel package (read-only mirror; do not push). Issue tracker for both Log Lens packages.

The LLVM of design — a JSON IR + MCP server that lets AI agents create KiCad, FreeCAD & glTF files.

The fastest way to build business applications. TypeScript framework with native AI app builder.

Portable memory layer for coding agents — persistent user & project context across sessions, installed into Claude Code.

Spec-driven development for AI coding agents. Seven core skills and optional domain extensions, from exploration to verified delivery.

A command center for coding agents you run on your own VPS — Claude Code, Codex, Antigravity & OpenCode in one cockpit, self-hosted, no Docker required

Mission Control for cost-conscious OpenClaw enthusiasts. Lightweight CLI/TUI tools for agent cost monitoring, task management, and configuration.

A 686-lesson library (9 domains, 54 subjects) authored by an AI coding agent, paired with Socratic discussion sessions to test and record what you actually understood.

Local multi-provider CLI orchestrator: a plain-text task list to LLM-planned JSON DAG to parallel dispatch across local agent CLIs.

Skills, plugins, and hooks that change how coding agents work — Claude Code, Codex, OpenCode, Antigravity. Measured, not vibed.

A lightweight Android AI chat client with multi-account and session management.

The local-first brain for coding agents: project memory + workflow intelligence + lossless ~50% context savings. MCP server + LLM proxy, zero cloud.

Mission control for your coding agents — every Claude Code and Codex session on your Mac in one window: live terminals, activity status (which agent needs you), restore after reboot.

A local visual studio for coding agents: author scenes, combine generated and recorded media, and export images, diagrams, animation, and video from retained sources.

Your AI says done. Verify real test, build and Git evidence with a reproducible trust score. MIT; built on agent-done-or-not.

Leave notes in the margins of your repo's docs then let your AI agent apply them, resolve the threads, and journal what changed & why.

iMBrace Community Edition is a self-hosted, open-source AI Operating System designed to keep your company's knowledge strictly on your infrastructure. Build context-aware AI agents, stateful DAG workflows, and native MCP tool connections with 100% data sovereignty.

在 DeepSeek Harness 里孵一只数码兽：吃你的真实工作长大，数码宝贝式分支进化，零 token、模型不知道它存在 | Digimon-style raising game for DSH — feeds on your real work, zero tokens.

Git-backed, conflict-safe memory authority for AI agents via MCP (Model Context Protocol).

Trade Stories DSL

Three small, independent guardrails for AI coding agents: an anti-AI-slop writing checklist, a code-comment hygiene checklist, and a standing multi-agent orchestration recipe hook.

고개만 끄덕이는 자문단은 잊으세요. 25인 거장 자문단이 당신의 어려운 결정을 로스팅합니다. AI ROASTING 자문단 · Claude Code 스킬

Local control plane for Claude Code: see every session, approve permission prompts from one queue, delegate tasks to headless agents in git worktrees — plus a plugin marketplace that ships one agent framework to every project.

A systematic multi-agent study of agentic engineering: field map, 8 sub-reports, and learning resources (July 2026)

AI 驱动的智能客服呼叫中心系统，让每一通电话都被智慧对待

Analyze any repository into agent-ready onboarding context: commands, entrypoints, security audit, quickstarts, and SKILL.md.

Layered, replicated, self-adjudicating code review orchestrator

Capture, replay, diff & audit AI agent and model runs as portable, signed evidence capsules. Open-source and self-hosted — runs in your own infrastructure, from a laptop to a cluster. No SaaS, no telemetry.

Generate source-of-truth docs from a codebase, API specs, or document folders. Get an indexed, version-pinned pack, hashed to its source. Then code or discuss with your Codex/Claude Code agent.

Headless CMS - Built for Agents. Governed by Humans.

MailWizz delivery server extension for Sendmux — route emails through multiple SMTP providers (Gmail, Outlook, custom SMTP) with smart rate limiting, automatic failover, and bounce/complaint webhook handling.

The agent OS. Host AI agents 24/7 across Hermes and Deer Flow. 🦌🪽