Custom agents

Production-grade engineering skills and specialist agent personas for AI coding assistants. Orchestrates the full SDLC, from spec to ship, with structured verification gates, anti-rationalization checks, and senior-level engineering discipline.

Myco — the mycelial layer for your AI agents. A macOS menu-bar app that shares one skill across every agent (Claude Code, Codex, Cursor, Gemini CLI, Cline) and unifies their chat history.

Model Context Protocol server for Destiny 2 API integration. Provides player lookup, activity tracking, clan rosters, and item/perk resolution via local manifest cache. Built with TypeScript for Claude Desktop and other MCP clients.

WhatsApp for AI agents — an MCP server powered by Zaileys. 60+ tools with progressive disclosure. Works with Claude Desktop, Cursor, any MCP client.

HMAC-signed boundaries that keep untrusted LLM content as data, not authority

MCP servers exposing data and tools to Claude, Codex and other MCP clients.

AI-ready repository intelligence engine for Python & Node.js. Analyze codebases, detect frameworks, build project maps, and generate LLM-optimized repository context.

PGAutoPilot is a production-ready PostgreSQL MCP server that lets any AI assistant safely interact with PostgreSQL using natural language. It combines schema-aware query generation, enterprise safety controls, and a minimal-configuration setup into a single executable.

MCP server for Bareos backup system. Query your backups via your favorite AI assistant

Token reduction & persistent memory for Pi coding agent.

Local, permissioned MCP server giving AI coding agents real Android dev tools — Gradle, adb, logcat, lint, crash triage — instead of raw shell access

A workbench for running large Mixture-of-Experts LLMs locally on consumer hardware with a tight VRAM budget.

MCP server over LS Securities OpenAPI — quotes, charts, screeners, ETF, program flow, and portfolio data for Korean and US markets, accessed in natural language.

Tool-agnostic agent workflow starter kit. One workflow core + team profile, generates adapters for Claude, Codex, Cursor, Copilot, CodeBuddy, Kiro, Trae.

A funky CLI that talks to Qwen 3.6 via Ollama with animations and attitude.

A local-first, file-based memory CLI for AI tools. Own your context in plain Markdown — auditable, versionable, and provider-agnostic.

🛰️ Talk to any AI agent runtime through one TypeScript client. Provider-agnostic, ESM, zero-dependency.

An Aesthetic Engineering Skill for AI Agents (Antigravity, Claude Code, Codex). Gives your agent visual judgment to eliminate "AI Design Slop", generate Design DNA, and enforce human-grade typography, layout, and spacing.

MCP client built from scratch in Python — STDIO transport, tool invocation, resource templates & prompt rendering via interactive CLI

MyAnimeList MCP server — Tenrai-powered reads plus official MAL personal-list management. Works with any MCP client.

Agent-readable semantic browser layer for AI agents — token-efficient web understanding, semantic snapshots, persistent sessions, and MCP/CLI/Python APIs.

Roslyn-powered MCP server for .NET — your coding agent stops reading whole C# files. 13-30x fewer tokens: semantic navigation, compile-gated edits, solution-wide rename, and type-checked XAML bindings for WPF/Avalonia/WinUI/MAUI. No IDE, no licence, one install.

Collection of Agent Skills for Claude Code, Cursor, Codex, Windsurf, and any agent that speaks the open Skills spec. Ships via npm + plugin + skills add.

Menu-bar app: PDFs, Office & images → clean Markdown on your Mac. Offline, local, no API keys.

Connect AI agents to HP ALM via the Model Context Protocol (MCP)

Local MCP code-intelligence for AI coding tools: surgical search/read/edit, agent-config version history, path-level access + masking guards, and a multi-project hub.

MCP server for .NET runtime diagnostics — memory, GC, threads, and more

A headless, filesystem-backed MCP server that serves any MasterVault — its files and its operating protocol — to any MCP client over stdio. No Obsidian required.

Open-source Agent + SDK for the xcircl verified regulated-care provider data API. Source-linked, timestamped.

AI Skill Finder 2026: Automated Search & Install for Claude Code

Local-first Chrome extension: inject selected web content into ChatGPT with reusable prompt templates.

Remote cockpit for CLI coding agents — drive Claude Code / Codex / Reasonix from your browser or phone. CLI Agent 网页远程座舱：浏览器/手机远程驾驶多个 CLI 编程智能体。

Principal-level Hermes Agent profile and skill pack for production RAG, AI agents, evaluation, observability, and implementation-ready architecture docs.

Security-first, brain-agnostic self-hosted agent runtime

Multi-stage LLM equity research agent that anchors deterministic bytes32 commitments of its trading decisions on Mantle for post-hoc verification, executes paper trades on Alpaca, and enforces a fractional-Kelly risk kernel. Built for the DoraHacks Turing Test Hackathon 2026.

AI Task Orchestrator • Parallel execution • TDD/PDCA/Reflexion loops • Methodology-grounded • Test quality audit • Merge gate verification

LEASH-8: an 8-domain control model for AI agents with delegated authority. Scorecard, approval-design checklist, plan-vs-authorize pattern. Patterns from a live production agent operation, sanitized. Free, MIT.

AI "company brain" for a pasta maker (Al Dente S.r.l.): a FastAPI bounded tool-calling agent over live CRM/ERP/call APIs + a compiled RAG wiki, with a Next.js/React UI featuring a 3D knowledge graph. Frozen POST /ask contract, honest abstention over hallucination.

Build one shared, permission-aware, auditable AI memory for an entire organization. A vendor-neutral blueprint you hand to a coding agent: 13 pillars, 70 decision forks, an M0-M15 build order. The org-scale sibling of RAG-OS and an alternative to Glean, Dust, M365 Copilot, Letta, and Zep.

The open Company State Runtime — version control for your project's state. Goals, decisions, work, and evidence outlive every chat, context window, and agent session. Ships the self CLI.

Decision layer samples for diagnostic reasoning and RoboCasa scenes for LLM agent evaluation

Run large language models locally on Windows. A native Rust desktop app and local LLM API server for GGUF models fully offline, no Python, no Docker, no cloud account. It measures your GPU and the model file, then loads a configuration that fits. Tuned on an 8 GB laptop card.

Visual editor and runtime for explicit agent workflows built on OpenCode.

Personal autonomous AI agent runtime in Rust. Multi-channel (Telegram, terminal, HTTP), background task handling, persistent memory, tool orchestration.

An event-sourced runtime for AI agents that survives kill -9, written in Rust.

Source-of-truth marketplace + project scaffolder for custom Claude/agent skills, personas, and instructions. Organized as {stack|common}/{feature}; welcome-ai-badger scaffolds a repo, feed-badger contributes back.

EVIDIQ Operator — Computer Use Infrastructure for Autonomous AI Agents. Browser execution MCP via 0G Compute TEE (GLM-5.2) + x402 USDT0 on X Layer.

Generic, language-agnostic AI-assisted development workflow for OpenCode — pipeline, agents, standards, and scripts

Personel Assistant On Telegram

A LangGraph superset: a 3D capability-manifold agent framework that routes over a semantic space instead of wired edges.

Obsidian + Claude Code vault methodology — agents, slash commands, wiki architecture. CC BY-NC 4.0.

A locally running RAG chatbot with safety guardrails and RAGAS evaluation, reproduced from three upstream repositories on a single workstation with **rootless podman** and **Ollama on the host** — no OpenShift, no Kubeflow, no cluster.

🧠 Don't Miss A Face — Automatically backup WhatsApp photos of your loved ones to Google Photos using AI face recognition

Stops your coding agent from pushing unverified work. Your agent does the reviewing; a state machine records the result per-commit and gates the push.

An automated, agentic exploratory testing tool that performs comprehensive QA testing on web applications, simulating human user interactions through various input methods (mouse, keyboard, TAB navigation, arrow keys, etc.).

See which Claude Code session is waiting on you — a macOS menu bar app showing your Claude Code sessions (local + remote over SSH), grouped by repo, surfacing the one stuck at a permission prompt.

TOKN — download & try. A trust-first agentic coding/reasoning harness with 20+ regulated-domain plugs, offline licensing, and in-place self-update. Public binaries only (source is private). Start a 14-day trial: tokn license trial

A case-routed knowledge base of software best practices and edge cases, written for LLM coding agents to load as precise, minimal working context (Karpathy LLM-wiki pattern) — routed by domain, one case per page.

Run multiple AI coding agents (Claude Code, Codex, Gemini) side by side — each in its own git worktree, all from one CLI. Web UI is just a window into it.

A unified local AI workspace — chat with GGUF models (Qwen, Llama, Phi), generate images, and share your AI remotely. Features agentic web search.

Terminal picker for Claude Code, Codex and OpenCode: harness, provider, model, system prompts, launch.

AI desktop guide that shows you where to click — it observes, never clicks. Free tier (no sign-up), BYOK or local Ollama, source-available.

EmotiWeave: correctable multimodal emotion recognition with continuous affect tracking and conflict detection.

A test-driven development loop for agentic coding tools such as Claude Code and Codex: /spec, /plan, /implement, with independent review agents that gate the plan before you code and the diff before you commit.

Systema Auxilium is an operating-system AI assistant that can handle a wide variety of tasks for you.

A method and toolchain for biblical study with an AI assistant — without ending up trusting it. Atomic cited findings, firmness marking, and a validator that keeps the corpus honest.

Native macOS desktop client for reviewing GitHub pull requests — PR dashboard, AI review drafts, CI diagnostics, and merge controls in one window.

TypeScript/Node.js MCP server for LinkedIn API integration - Create posts, manage content, and interact with LinkedIn through Claude Desktop and Windsurf IDE

OpenClaw AI Assistant 的核心文档与配置仓库

OpenClaw channel plugin — Telegram userbot via MTProto (GramJS). Machine-bound session encryption, DM & group support.

政企会议纪要skill

🎯 AI agent skill for optimizing Next.js applications for search engines

Chinese interview-coach skill for AI-assisted job search, structured practice, answer critique, and interview prep workflows.

lightweight, high-performance Personal AI Agent assistant workspace built using Node.js

Native Windows meeting overlay with real-time transcription and local-or-cloud AI answers. Pure Rust + Slint.

Pebble Time 2 remote for monitoring T3 Code agents, reading responses, and dictating prompts from your wrist.

8 款生产级 AI Agent Skill：持久记忆、代码安全评审、自然语言数据分析、调试复现、测试生成、IaC 安全扫描、性能诊断、文档生成 | 8 production-ready agent skills (memory, code-review security, NL data analysis, agentic debugger, testforge, iac-guard, perf-doctor, docsmith). Pure Python stdlib, zero dependencies.

AstrBot 插件：SSH 远程接管 Linux（多服务器）；ssh_exec / send_file / save_file；仅 QQ；白名单 + 致命命令闸门

CMM: Correcteur Mathématique Markdown

Operate your household finances instead of winging it. A local-first Claude Code skill pack + templates for households.

MCP server — 97–99% HTML token reduction via DOM AST extraction. No API key, no Vector DB.

Offline, single-file rubric for validating AI outputs — routes the right checks, writes the prompt, catches the stale number

MCP gateway federating 21 biomedical MCP servers — gnomAD, ClinVar, HPO, UniProt, Ensembl VEP, PanelApp and more — behind one Streamable-HTTP endpoint, with collision-free namespaced tools and BM25 tool search.

Enable AI agents to control Windows tasks using PowerShell skills for Outlook, Edge, desktop automation, and structured shell commands in JSON format.

Multi-mailbox IMAP/SMTP control plane. One YAML, N accounts, one bearer-auth gate. Unified inbox across Gmail/Outlook/iCloud/mailo/inbox.lv/whatever — search, read, mark, send. Reader mode strips HTML chrome for LLMs. Streamable-HTTP MCP server + REST API on one port. Ships as a Docker image.

Manage your career profile in Obsidian with a modular vault for CVs, applications, and professional history

Requirements discovery as an Agent role: Scout-driven dialogue turns vague ideas into implementable, verifiable requirements - Anchorlaw's upstream input contract.

Autonomous Bybit spot trading agent: an LLM proposes orders, a deterministic risk guard validates every one before it reaches the exchange. Safe by default (testnet + dry-run). No proven edge — read the measured results.

A multi-agent Python system for comprehensive stock analysis and equity research reporting

🛡️ Modular Python framework for Cybersecurity LLM agents, purple team automation, log triage, and detection engineering in lab environments.

Can AI agents discover that the laws of their world have changed? An instrumented artificial world where LLM scientist societies run experiments, exchange letters, and form beliefs under hidden ground truth — seeded batteries, frozen manifests, and provenance-checked evaluation. 235 runs across two studies; combined paper under review at JASSS.

Open-source multi-agent simulation engine to predict society-level response to launches, crises, campaigns, and narratives.

Reference implementation of Double Ratchet: co-evolving an inspectable evaluation metric with a lifecycle-managed skill library for self-improving LLM agents (arXiv:2607.12790)

Open-source infrastructure and an emerging specification for traceable, revocable, rebuildable, and verifiable persistent learning in AI agents.

A cognitive operating system for AI agents — browser & computer intelligence, persistent missions, memory, recovery, evidence, and governed real-world action.

FUTS program evolution: evolve a program toward the best score on your metric — 以coding agent为变异算子的程序进化框架

Local LLM agent that operates a VyOS network lab over SSH. Tool-calling, evals, and fine-tuning on a 3B model running fully local.

Self-hosted 3D orchestrator for multi-account Claude Code agent teams: rooms as departments, a factory message bus, an orchestrator agent, and per-subscription rate-limit monitoring.

Nine-agent AI framework automating quant research from hypothesis generation through leakage-aware backtesting

Defense-in-depth against prompt injection in tool-using LLM agents. Semantic drift detection catches synonym-evading attacks that bypass regex. Self-play red-team, calibrated scoring, ablation study.