Code agents

Local-first Python SDK for AI agents: Shadow DOM flatten, Action Map, SchemaSnap, form fill, security scan, MCP (64–97% token reduction)

A Next.js starter showing how to run Claude Code / Codex in safe, verified loops — plan → build → report cycles with lock files, budgets, denylists, and a human merge gate baked in.

Control Spotify from your AI Agent, easily and securely

A curated list of 100+ Model Context Protocol (MCP) developer tools — SDKs for 15 languages, frameworks, inspectors, testing and security scanners, proxies, gateways, hosting, and templates for building MCP servers and clients.

Your AI agent develops firmware on its own: build, flash, stimulate real hardware, observe, fix - policy-gated MCP tools + pytest plugin.

Living Docs — a AI agent skill that runs your project's documentation as a living system: constitution, ADRs, BDRs, PRDs, issues, research notes & Mermaid architecture diagrams under five no-drift governance invariants. OKF-formatted, stack-agnostic. Works with Claude Code, Cursor, Copilot, OpenCode & Pi.

“What’s the thing?” You’d answer: “Skills that teach you how to maintain skills & repos”. Engineering Stewardship for any repository (app, library, tool), a structural governance and quality layer for agent-operated engineering repos.

Safety and contract checks for AI-generated n8n workflows

Self-hosted MCP server + dashboard giving Claude Code, Cursor & Hermes agents shared long-term memory over a git-backed markdown vault. Agent-native, Obsidian-compatible second brain — markdown RAG, no vector DB.

Codex skill that disables Responses Lite model mapping to restore hosted Web Search through OpenAI-compatible gateways.

Persian-first, machine-readable directory of free LLM APIs with quotas, OpenAI compatibility, official sources, and Iran-access evidence.

Queue research questions all day. Wake up to answers. Puts your idle Claude Code quota to work while you sleep.

Replace the vendor app for your Pocket AI voice recorder. Pull recordings over BLE, transcribe locally with Whisper, diarize speakers, and summarize with your own API keys — no vendor cloud, ~$2/year.

LLMForge is probably the most modern AI vulnerability lab for OWASP VulnerableApp — real-LLM-backed labs for prompt injection, LLM-orchestrated BOLA, and RAG attacks.

Open-source Python toolkit + playground for Vietnamese AI — RAG over docs, diacritic restore, segmentation, OCR, normalization. Local-first, multi-backend (Ollama / llama.cpp / HF Transformers / OpenAI / Anthropic).

aprovodka — MCP server for 1C:Enterprise via REST/OData 3.0. 32 tools: catalogs, documents, registers, accounting, batch, change-tracking. No BSL extension required.

Hire a 42-employee AI company in one command: 7 departments, 42 skills, 1 CEO. For Claude Code & any AI CLI.

Desktop app to test, inspect, and debug ACP (Agent Client Protocol) agents — a JSON-RPC protocol inspector, like MCP Inspector for ACP.

Your coding agent, wired to 4,000 developers' real-world knowledge. Ask via /rombot-cli — grounded, cited answers, not web-search fluff. Claude Code, Codex, Cursor, Pi.

Curated AI tools, Claude skills, MCP servers & growth playbooks for solopreneurs, indie hackers and solo founders — the one-person-company AI stack you feed to Claude Code.

Agentic AI monitoring and observability — OTEL traces and local log files from Copilot, Claude Code, Codex, and OpenCode. Token usage, cost, summaries, and efficiency insights. VS Code extension, npx, and Docker.

🤖 The strongest curated list of open-source AI-agent tools for literature review — 70+ tools across 11 categories (Claude Code skills, deep research, MCP servers, PDF parsing, systematic review). 面向智能体文献综述的最强工具大全。

Lightweight, agent-friendly AI security toolkit for Node.js & TypeScript. Stop prompt injection, audio hallucinations, and RAG data scraping. Zero dependencies. MIT

Consort keeps AI-written code clean and correct: a coordinated ensemble of role agents, spec-first and test-driven, driven by a deterministic state machine against live branches of a real Lakebase database. Engineering discipline is no longer left to the whim of a model. Consort enforces it.

一个现代化、跨平台的AI智能终端，支持SSH、SFTP、端口转发，通过与AI对话运维服务器，让小白也能成为运维高手

AI customer support agent with semantic search — handles Gmail, WhatsApp & web via OpenAI Agents SDK + FastAPI

Companion code for the Manning book 'LLM Customization and Fine-Tuning.' Adapt open-weights LLMs end to end: prompting and RAG, LoRA/QLoRA, full SFT, distillation, and DPO/RLHF alignment.

내 맥에 상주하는 AI 팀. 텔레그램·슬랙에서 일을 맡기면 누가 담당인지 정하고, 진행이 대시보드에 남습니다.

AI compliance platform that uses RAG to automatically match AI systems to state regulations, generate grounded disclosure documents, and maintain a cryptographically-signed tamper-evident audit trail.

1282 production-ready AI agent skills with self-evolving meta-skills — 783 cybersecurity, SEO, trading, marketing, automation, devops, content, financial. Auto-evolve, auto-learn, auto-improve. Install: npx skills add oyi77/1ai-skills

Claude Code experience for the pi coding agent, in one package: reads your .claude config (rules, commands, skills, hooks, output styles, MCP, agents) and adds todo, checkpoints, memory, web, and subagents.

Run Codex and Claude Code reliably in the background on macOS—with a native dashboard, persistent tmux sessions, local checkpoints, recovery, notifications, and closed-lid support.

Portable, self-contained sandboxes for AI coding agents - microVMs, containers, isolation harnesses.

DyroEngineeringFlow 是面向多仓团队的本地优先工程自动化与交付控制平台。它将开发线、Git worktree、任务编排、 Agent 启动、质量门禁、独立复核与合并审计统一到可版本化配置中，帮助团队从任务创建到交付合并形成清晰、安全、 可追溯的自动化流程。日常通过简洁命令 dyro 即可完成初始化、环境检查、任务执行与协作交付。

Evolve LLM prompts with a genetic algorithm: score candidates on your test set across real models, breed the winners, keep the champion that survives a holdout. CLI + Electron desktop app.

Sol + Luna layered subagent templates, skill, and demo for Codex / Claude Code / Pi (high-IQ leader + cheap workers)

AGENTIC · LEARNING — Hand this repo to your agent. It interviews you, hunts down the best sources in the world on your subject, and builds you a learning dashboard. Zero shared keys.

Evidence-driven AI software synthesis and verifiable artifact delivery platform with stateful orchestration, layered memory, hybrid retrieval, build validation, and reproducible deployment.

Agent memory that needs no API key, no LLM, and no embedding provider. One command on Bun and it is serving MCP over stdio against a local SQLite store. Drop-in for the MCP reference memory server. Every memory keeps its source, who may read it, and the evidence that contradicts it.

Learn Codex

A bwrap+nix sandbox with cool tools like opencode, jcode for AI agents experiments

Skills for working with https://www.mossnotes.app/

Run cyber-AI agents against real missions in isolated environments. Record everything as OpenTelemetry evidence. Grade the evidence, not the claim.

Local MCP switchboard for Codex, Claude Code, Gemini, Antigravity & VS Code — route tasks, run cross-model debates, and share context using the subscriptions you already have. No API keys, no cloud.

Turn an economics paper into a seminar-ready Beamer talk with a timed speaker script — conference, job-market and discussant slides. Agent Skill for Claude Code & Codex.

与 AI 长期协作的工程方法论 —— 每条规则背后，都是一次真实事故 | Battle-tested rules for long-running AI-agent collaboration, forged from real production incidents

一键发送 ChatGPT Codex 邀请的独立网页工具，支持 Vercel 一键部署 / Send ChatGPT Codex invites with one click

Any agent that can run curl gets a real browser. Snapshot a page as refs, click one, read the JSON behind it — no SDK, no MCP, no API key.

A syscall-level security supervisor for AI agents.

A Claude Code skill that turns Claude into a vSphere administrator using govc. Inventory, reports, snapshot audits, root-cause analysis, and safe VM/host/cluster management for VMware vCenter and ESXi — read-only first, with confirmation guardrails for destructive operations.

Compose neural architectures as typed, executable graph cards with two-way PyTorch sync, step-by-step local execution, and a constrained AI graph planner.

Navin - free local AI workspace for agentic coding, scraping, notes, meetings, SEO, leads, marketing & media. Windows · macOS · Linux.

The terminal your agent team lives in — run coding agents in panes on a shared bus, on a fast GPU-rendered terminal for macOS and Linux.

A hand-drawn illustration and animation language for LLMs

CLI, MCP and Skills for the world's property portals. Search, deduplicate and monitor listings across 9 providers in 9 countries — UK, Ireland, US, Canada, Germany, Australia, Spain, Italy, Portugal. Built for AI agents.

A cognitive memory architecture for agents: ACT-R activation, association, consolidation, and forgetting. sqlite, zero dependencies.

Your YouTube channel, encoded. A team of agents running ideation, hooks, retention, scripts, thumbnails, and distribution. Runtime agnostic.

Agent skills that turn your scientific reading and research history into persistent LLM-maintained markdown wikis, not one-shot RAG. For Claude Code + Codex CLI

Local-first action assistant for durable, reviewable software work.

Quality-first Codex skill for verified delegation across native subagents and Codex/Claude CLI workers.

YunkunSec（云鲲安全）— LLM-first 全栈式自动化网络安全智能体平台，集成 Agent Loop、Skills/RAG、工具编排、证据链、漏洞验证与报告生成。

Antigravity Observability Console & Wire-Tap Module

Skill para Claude Code — Sistema de 6 agentes IA para criacao completa de produtos digitais low ticket. Pesquisa de mercado, copy, VSL, pagina de vendas, criativos e estrategia de Facebook Ads.

FAAAH (Filesystem As An AI Handler) - Reuse your AI Agent subscription via text files.

Documentation-discovery telemetry for Claude Code — heat/cold maps, health grade, evidence-backed router fixes. 100% local, zero tokens. /tt

Oh My Pi host plugin for the GSD Embeddable Orchestration System — extension bindings, slash-command surface, and Host-Integration SDK adapter for AI coding agents

Seedance Drama Maker — turn one sentence into a cinematic AI short drama. Multi-agent pipeline powered by SeeDance 2.5 (30s shots, 50 reference inputs, 14 languages, native audio-video sync), SeeDream 5.0 Pro image generation, FastAPI + WebSocket.

Local, free voice artifacts for AI agents — CLI, localhost API, and Agent Skills.

AI 面试训练系统：跨会话能力建模闭环 — LangGraph 编排 + 三层上下文融合出题 + Mem0 式画像更新 + RAG 知识增强 + SM-2 复习调度。FastAPI · React 19 · LlamaIndex · Qdrant

Native Android GUI for running AI coding agents locally on-device. No terminal or PC required.

A fun Claude Code CLI with animated themes — prompt it, and it builds you your own!

Open-source alternative to LangGraph Platform for TypeScript — self-host your LangGraph.js agents behind the standard Agent Protocol API. Your infra, your Postgres, no vendor lock-in. Drop-in for the LangGraph CLI.

Welcome to the AI-Powered Workout Application, a cutting-edge fitness platform that empowers users to generate personalized workout plans based on their goals, preferences, and convenience. Harnessing the power of artificial intelligence and modern technologies, our platform offers a seamless and intelligent fitness experience.

Codebase Refactoring Agent Skill. Scans source files, runs standard audits, and proposes drop-in code refactoring.

Jetson Nano edge AI vision Agent Harness: real-time video analytics, MCP tools, event evidence, DeepSeek online/offline, secure recovery

Diagnose Chrome Manifest V3 extension pitfalls — and teach AI coding agents to avoid them.

Karpathy-inspired LLM Wiki demo: AI agents compile raw sources into a persistent, self-linking Markdown knowledge base. The Agent writes. The Wiki remembers. Live interactive walkthrough + local Streamlit app (Gemini).

Source-free architecture maps and deterministic polyglot refactoring for AI coding agents

常驻 Windows 顶部的 AI 编码 Agent 监控/审批/协作面板 + 个人工作台（Electron + React）

A Skill and CLI tool for Codex and Claude Code that converts images, PDFs, image-based PPTX files, and mixed PPTX files into editable PowerPoint presentations.

TikTok marketing skills for Claude Code and Codex: 3-second video hooks, captions, trend mapping, and a weekly plan. Draft, humanize, and publish via Publora. MIT. Content engineering by Creative Content Crafts.

A cross-platform desktop multi-project agent manager for the AI coding era, with integrated Agent sessions, Git management, and IDE launch. 为 AI 编程时代打造的跨平台桌面多项目Agent管理器，集成 Agent 会话、Git 管理与 IDE 启动。

Give the agent a cage, not your keys. One-command Docker sandbox for AI coding agents: full autonomous permissions, per-project isolation, your host stays untouched.

Local code-intelligence pipeline for AI coding agents: architecture maps, hotspots, change impact, and structural regression gates before the agent edits. Rust core, powered by rg + Repowise + Understand + Sentrux.

Six Claude Code skills (architect, review-swarm, ask-the-council, prompt-generator, up-to-date, ponytail) bundled as an installable plugin.

LidRun keeps your Mac awake with the lid closed — so Claude Code, Cursor, Docker and Ollama keep running with timers, charging-only mode, low-battery stop and battery/thermal guardrails.

Free desktop app: run Claude Code with multiple accounts and auto-switch when one hits its 5-hour limit. 24/7 auto-loop, parallel chats, live cooldown timers. No API key needed.

Package manager for AI agent assets — 42 skills, 5 CLIs, 1 plugin. Skills for AI/LLM agent engineering, eval-driven dev, spec-driven dev, database (MySQL/PG/Redis/Kafka/Mongo/Oracle/NL2SQL), K8s/Docker diagnostics, infra ops (Dify/Nexus/Harbor/SonarQube) & docs/patent. Runtimes: Codex, Claude Code, Cursor, Kiro, DSH.

Automatic conversation between 2 OpenAI GPT powered characters who participate in a Turing test together.

专为sub2api打造的iframe嵌入式AI图片生成工具，纯前端实现，无依赖开箱即用。支持文生图、图生图、多种分辨率、批量生成和历史记录功能。

Track your daily prompt contributions across CLI coding agents

Your agent can read git log. It can't read the shell commands you tried that failed. NexusMem indexes shell history with exit codes, git commits and diffs, and project docs into a local SQLite database, and serves back a token-budgeted slice on demand. No cloud, no account.

From vibe coding to knowing what you actually wanted. Point it at a project you already built with an AI agent: it reads the conversation back and derives the real requirement from your own corrections, layers the mistakes, and turns repeated work into a Skill. Counting is mechanical. Claude Code + Codex, EN/中文.

Bring Trustabl's AI-agent safety & reliability scanner into Cursor — scan your agents, tools, and MCP servers, get a production-readiness score with severity-ranked findings, and fix issues before they ship.

Import pi / opencode / codex / claude-code sessions, chat history, and agents into DeepSeek Harness — one-click Sync button, slash commands, session-start migration prompt

Cortex is a sophisticated AI coding assistant that combines the power of multi-agent orchestration with Retrieval-Augmented Generation (RAG) to provide intelligent, context-aware answers about your codebase. Built with privacy in mind, Cortex runs entirely locally using Ollama or can be configured to use OpenAI's models.

让 AI 像人说话的完整规则集。去除 AI 腔、讲义腔、协作痕迹。适用于 Claude、ChatGPT 等所有对话 AI。 EN：: Complete ruleset to make AI talk like humans. Remove AI-speak, lecture tone, and collaboration traces. Works with Claude, ChatGPT, and all conversational AI.

Use your OpenAI subscription with DeepSeek Harness to access GPT models, image generation, and web search.

Machining process planning for motor shafts - input shaft geometry, RAG-grounded operations, machine tools, cutting tools, heat treatment, exported as a process-card Excel.

Auto-generate AI assistant rules (CLAUDE.md, .cursorrules, copilot-instructions) from codebase analysis