Trending Repos

Open-source dots for the web: an AI agent with its own browser, one that does not get blocked.

AI生成の日本語を自然な日本語へ推敲するAgent Skill / Agent Skill for Refining AI-Generated Japanese into Natural Japanese

UI design skill for Claude Code, Codex and other AI agents: explore distinct directions, compare them side by side, refine against real screenshots. 给 AI Agent 用的 UI 设计 Skill。

Token 超支？快给史密斯打电话！ Better Call Smith. Adaptive model routing, a glassy Codex usage monitor, and reproducible benchmarks with receipts.

Astra plans and reviews; DeepSeek Flash builds. A native Codex workflow with phased tasks, verification, safe installation and reversible setup.

CodeAgent is an open-source coding agent that lives in your terminal. It understands your repository, plans multi-step changes, writes and edits files, runs commands and tests, and iterates until the task is done — all with a transparent, extensible tool loop.

Sprix AI at 屿智同行 × 保研岛 — an open-source benchmark for AI agents in Baoyan advising and application planning.

AgentJev-0.6B - a fast 'System One' decision model for AI Agents: feed it any unstructured state (diffs, traces, logs) and structured questions, get calibrated probability distributions back in one ~50ms forward pass. Zero output-token decoding.

Track and visualize the stars history of any GitHub repository. Open-source growth analytics and velocity tracking.

Turn reference videos into editable motion graphics with Codex + ChatGPT. Compare, customize, and export MP4s or HyperFrames projects. Local Windows app + online studio.

A self-hosted coding agent for background work - works with Claude Code, Codex, OpenCode & 100+ providers

FastAPI service for Stellar account screening, Soroban event monitoring, and network health.

VoiceShell（声壳） · Experimental Muse bridge: ESP32 pairing, Linux text round trips, real walkthroughs and compatibility notes.

Benchmarking Coding Agents for Interactive World Modeling from Embodied Videos - Rethinking Embodied Real2Sim from a Software Engineering Perspective

Jev demos, projects, SDKs and skills, with source links and a curated X gallery.

Native, non-invasive parallel multi-account & sandbox manager for Google Antigravity

共享智能体，共享项目上下文，让项目所有成员一起协同工作。同时智能体会像人一样通过资料或者工作抽象和总结经验到项目上下文中

Audit a GitHub repository before launch - ten readiness checks scored 0-10: name, description, topics, README, license, demo, installation, contributing guide and issue templates. Free and open source.

Free open-source extractor for AI coding assistant chat histories. Supports Claude Code, Cursor, Windsurf, Aider, Cline/Roo Code, and more.

Audit your GitHub repository search ranking signals. Checks name, description, topics, README, stars, forks, and activity.

An open research workbench for idea discovery, autonomous mathematics, and evidence-grounded rebuttals.

System 1 decision models (Jev, Laya, Cua-S1) as brain for agents: Browser use, computer use, games and robotics

CoreML-native local decision runtime for AI agents · 18 install profiles · 20-scenario protocol · offline · ~50.93 MB · by Soulite Magic

Audit AI agent tool-call transcripts: hash-chain sealing, deterministic replay, and scope overreach checks. Dependency-free Python CLI.

CODING — 一款以本地优先为核心的AI编程代理，支持基于代码仓库的规划、执行、验证与交付。

A terminal coding agent that plans work as executable graphs

PaperGraph MCP turns math papers into evidence-grounded reading maps for AI agents: extract results, trace proof evidence, plan reading order, and review external dependencies without guessing.

FOREST is a research agent for long-horizon science, where evidence gets lost, errors compound, and goals drift. It explores ideas, reads literature, forms hypotheses, runs real experiments, analyzes results, revises paths, and builds a full paper—all in an editable graph you can inspect, branch, redirect, with claims and figures linked to evidence

STAMPEDE — read-only explorer of observed wallet rotations on Robinhood Chain

A curated collection of papers, research blogs, open-source tools, benchmarks, and community demos for robot-use agents, including demos powered by GPT-6 Astra. —— Explore our searchable website.

Yerel, çevrimdışı ve taşınabilir İngilizce pratik uygulaması (Türkçe konuşanlar için): yapay zekâyla sesli/yazılı sohbet ve anında düzeltme, A1–C1 dersler ve sınavlar, Türkçe sesli öğretmen, kelime ve kalıp çalışma, YDS tarzı süreli testler. Windows ve Linux. Ollama + Gemma 4, Whisper, Kokoro, EMA Lightning.

🎨 Agentic skill for Qwen-Image-2.1: Rewrites and optimizes text-to-image and multi-image editing prompts using official Alibaba specifications. Compatible with skills.sh and all AI agents.

High-performance OpenAI-compatible API gateway and bridge for muse.ai personal AI agents, ready for 9Router multi-model integration.

Established strong JEV based natural language to SQL performance and well-covered semantic operators for PostgreSQL(PG) with optimized cost and latency. Reusable evidence, with native execution, parallel plans and explainable embeddings in development preview.

Unofficial skill that teaches coding agents to build with TypeSafe AI's Jev: typed decisions, calibrated confidence, and prior art from 150+ community projects.

A small, complete Python agent harness: tools, compaction, retries, and a single-file TUI. Built to learn and use as a baseline.

Claude Code agent skills for game dev in Blender. Turn concept art into rigged, game-ready 3D assets, measured against your references.

👋 Hello Agent System — learn enterprise AI agent system design by building every production layer from scratch. 17 bilingual lessons (EN/中文) with tested exercises: tools, architectures, reliability, security, evals, distributed execution, cost, RAG, release ops. 企业级 Agent 系统设计训练营

像coding agent一样自动投递boss直聘

Personal AIOps engineering project with durable workflows, scoped hybrid retrieval, reliable event delivery, and verified API/browser contracts.

Local AI search for podcast and video archives on Apple Silicon: Whisper speech, Apple Vision OCR and SigLIP-2 visual search fused into one ranked list, with a trim editor and FCPXML export. FastAPI, vanilla JS, Tauri. Source-visible, all rights reserved.

Cinematic launch films and motion design from code, directed by your coding agent.

中国市场数据 / 中国股票市场数据 Skill：A股行情、K线、逐笔、研报、资金流、公告、ETF/期货/宏观。面向 Claude Code 与 AI Agent 的 Python CLI。China stock data skill for agents.

Who the good traders on Robinhood Chain are buying: resolved wallets, a 20-second on-chain tape, provenance on every fill, AI verdicts, bursts and exits. Site + Telegram bot + API. Runs on $0/month.

Agent evals and guardrails as Jev decisions: one request per trace, a fraction of a cent, fast enough for the agent loop. Runs locally with Kev or Laya.

Creative vertical video editing skill for AI agents (Claude Code, Codex, OpenCode, Cursor, Hermes): named styles, creative mode, HyperFrames + voice-first mix.

Multi-agent pipeline that turns a research topic into a draft paper grounded in real literature, with citation hallucination detection and claim-support checking.

Local-first recruitment intelligence and application operations agent

Agent2 — An autonomous AI agent for coding, cybersecurity, terminal automation, and intelligent task execution.

WMPFDebugger 内置浏览器与公众号 H5 满血调试与逆向工具箱 / The ultimate debugging & reverse-engineering toolkit for WMPFDebugger In-App Browser & H5 Webpages with Frida hook, vConsole proxy, and API analyzer.

An open-source multi-agent platform for GitLab MR code review, governance, and CI self-healing.

OpenTelemetry plugins for every coding agent, in the Agent Plugins format — a marketplace that runs itself: open an issue, the repository validates, reviews, lists and follows your plugin.

Bicameral deterministic AI runtime: Zero-hallucination System 1/2 reasoning engine with cryptographic audit trails (RFC 8785, Ed25519) and PoSP consensus arbitration.

Open decisions model with Jev's API: typed answers (yes/no, choice, score, multi) with probabilities from forward passes, no generation. Qwen3.5-4B + LoRA, one GPU.

AI humanizer skill for Claude Code, Codex and Cursor. Removes AI slop and the signs of AI writing, then measures the rewrite against human baselines.

Evidence-based code review skills for Codex and Claude Code

Behavior-guardrail hooks for Claude Code: test-tampering + outbound-action guards

Open-source FDE project workbench for research, AI opportunities, delivery and team collaboration.

Open-source marketing system, designed for coding agents.

基于 AI Agent 的产业园区项目管理研究 · 原型验证平台（Prototype，不代表最终产出产品）

Control Android phones from coding agents through MCP, a CLI, or a skill. No API keys required.

面向 AI Agent 的高并发、低延迟上下文治理与 Prompt Cache 守护网关 (Tree-sitter AST / Tool Delta / 50%~80% Token 削减)

SkillAdam: Better Skills for Your AI Agent🚀 Claude Code/Codex/Cursor 插件，一键进化你的skill

Local-first document requirements and evidence review workbench with traceable citations and human review.

Autonomous X/Twitter posting pipeline for your AI agent. No API key, no paid scrapers: researches your niche, writes in your voice, publishes through a real browser, and verifies its own posts.

Python AI Agent 学习计划（精简大纲 + 学习详情）

Slot continuity and fork analysis for captured Solana slot records - gaps, skips, duplicate slots and orphaned branches, with an independent Rust engine.

标书匠 BidCraft · 对话式标书编制系统：AI 智能编制技术标，支持招标文件解析、标段选择、标书编写、知识库提炼与学习对话式改稿。LLM-Wiki 路线，基于LangGraph实现。Open-source AI agent for construction tender & bid documents.

Novel to comic workflow and skills for Codex and AI coding assistants.

Argolink-focused review automation with a coding model for Claude Opus 5.

为 AI 打造的《魔兽争霸 III》1.27 开放接口：用任何语言写 AI · An open API for Warcraft III 1.27, built for AI agents — write bots in any language

Recoverable and auditable deep-research agent built with LangGraph. It plans research, searches the web, manages evidence and context, resumes interrupted runs, and generates cited reports—also designed as a learning reference for AI Agent enthusiasts.

Teaches AI coding agents to make clear, honest charts, and checks each one before you see it. For Claude Code, Codex, Cursor, and more.

A production-oriented multi-agent AI Agent system for software engineering. | 面向软件研发任务的多智能体 AI Agent 工程系统

Show it. Say it. Your AI gets it. Record your screen and talk — your coding agent turns it into bugs, ideas and to-dos, with the exact frames marked. 口喷鸡：边看边喷，AI 全懂。Agent Skill + macOS menu-bar app for Claude Code / Codex.

Pi-native revisable evidence dataflow for the GLM/Gemini/GLM-Flash matrix

Scriptable Cloudflare Turnstile token CLI via the Peak API - sitekey discovery, solve, and token output for CI and automation.

Repository retrieval and context engine for Claude Code and Codex: finds relevant code, reuses project knowledge, and pairs each task with specialist skills and verification.

Zero-LLM cognitive memory architecture for AI agents

A lightweight local mock LLM provider — deterministic OpenAI and Anthropic responses, with no API calls, no tokens, and no code changes.

Homework assignments for Evaluating and Improving AI Agents

冰箱水墨屏：基于 ESP32-S3、四色水墨屏、微信小程序和 FastAPI 的低功耗开源信息屏

AI 桌面换装视频怎么做？把全网刷屏的「桌面动态壁纸换装」逐帧拆解，整理成一条可直接复制的提示词模板：一条提示词直出 30 秒成片，固定机位一镜到底，中文对白和字幕全部由模型生成，无需后期剪辑。适配即梦、豆包、Seedance 2.5 等 AI 视频工具，附火山方舟 API 脚本，并封装成 Cursor / Claude Code 可一键安装的 Agent Skill：说一句想法，它反过来采访你，自动写出精准提示词。

Free local guardrail for AI coding agents: every tool_call judged on-device (regex + Laya), allow/review/block

A lightweight coding agent 🧑‍💻 built from scratch, with tool calling, session persistence, workspace context, memory, and multi-provider LLM support 🛠️.

Verify that an AI agent's on-chain claims are actually true. A claim-level groundedness checker for Solana / memecoin agents.

⚡ The Ultra-Lean AI & High-Performance Developer Linux Distribution. Sub-350MB idle RAM, dynamic zRAM (ZSTD), 45 native GTK3 zero-terminal GUI apps, and 1-click local LLMs.

Alice 集成技能包 · 放进去就能用 · Drop-in skill suite for AI coding agents

Ψηφιακή κάρτα εργασίας για το ΕΡΓΑΝΗ · Digital work card for Ergani

Gravitas is a free add-on for Google Antigravity that makes its AI agent check what your answer depends on, tell you kindly when something you assumed is wrong, and still give you what you came for.

AI agent skill for maintaining project AGENTS.md and docs/ documentation systems.

闲鱼多账号自动回复与经营管理工具 — Electron 桌面端 + FastAPI 后端。扫码登录、关键词/AI 回复、订单发货管理。

Delegated authority for AI agents. A Pigeon Pass says what an agent may do.

Bulk-provision TokenMix accounts with disposable temp-mail.io inboxes and auto-generate API keys with random names. Turnstile-aware, Playwright-driven, JSON/CSV output.

投镜 · Personal Investment Twin — an AI investment decision coach: rebuilds your investment episodes from your own trading history, attributes P&L with deterministic computation, and reminds you before you repeat past mistakes. No stock picks, no prediction, no auto-trading.

Keep the signal. Cut the noise. Local CLI + MCP — fold verbose tool output for Claude Code, Codex, Cursor & Desktop.

Apostate is a 100% free, fully open-source anti-detect browser built on Chromium. Defeats every major commercial fingerprint detector with zero paywalls and fully verifiable GitHub Actions builds. 153 patches. FPJS Pro: suspect score 0. BrowserScan: 100% authentic. deviceandbrowserinfo.com: human. bot.sannysoft.com: all green. With proof.

Grades the retrieval policies that decide what a long-running agent remembers - comparative scorecards over fixed, inspectable sessions

Vibe知识大赏风格 AI 知识视频 Skill — 纯代码生成电影感知识短视频（无配音·双语字幕·合成BGM·历史文献再现），风格规范逆向自 8 条抖音爆款原片（1091 帧逐秒分析）

Поэтапное ревью кода 1С AI-агентом по ключу задачи GitLab или GitLab Merge Request

Примеры кода на Python 2026: лучшие практики, разбор примеров, Python 3.14, asyncio, FastAPI, pytest, задачи с решениями на русском

Guia extenso sobre o Maestri: +257 templates de partitura em 12 áreas, documentação, dicas e recursos para reger times de agentes de IA. Para ajudar, agregar conhecimento e tirar dúvidas — tudo para quem trabalha com tecnologia.

Jev 决策模型驱动的 UI 自动化测试框架：不生成「下一步做什么」的文字，而是在候选元素里直接打分；一句话写用例，pytest 执行，Allure 出报告，每个操作决策中位 458 ms（UI test automation driven by the Jev decision model — scored decisions, not generated prose; ~458 ms per decision）

A pure-Python system that automatically discovers, verifies, and applies relevant historical requirements for coding agents — no embeddings, no vector database, no LLM calls. Companion code for the Towards Data Science article on intent continuity.

Evidence-driven Java security review workbench

Desktop virtual wind tunnel: drop in an STL, get OpenFOAM CFD (drag/lift, pressure maps, flow slices, streamlines)

Flag ATTACH DATABASE footguns — cross-database atomicity, alias shadowing, ambiguous unqualified tables.

OpenAI- and Anthropic-compatible API for Duck.ai, driven by a real Chrome browser via Playwright. Streaming chat, image generation, session reuse, and a web dashboard.

Desktop virtual wind tunnel: drop in an STL, get OpenFOAM CFD (drag/lift, pressure maps, flow slices, streamlines)

Universal skills and agents for every AI harness. Author once, wire anywhere: Claude Code, Cursor, Codex, Copilot, Gemini CLI, Windsurf, Cline, Aider.

Local-first Agent RL workbench for Search-R1 rollouts, reward evaluation, GRPO/verl training-data handoff, and an auxiliary private RAG workspace.

Real-time Steam game leaderboards (最热游玩 · 热销 · 特惠 · 新品 · 免费游戏 · 限时免费) — live player counts, CNY prices & deals, auto-refresh every 60s.

简明技术中文 / Simplified Technical Chinese: a controlled language for Chinese technical writing, modelled on ASD-STE100. 40 rules, a one-meaning-one-word vocabulary, a zero-dependency checker, and an agent skill for Claude Code, Codex and Cursor.

A fruit fly connectome (MaleCNS v1.0) trading an intraday NIFTY straddle through OpenAlgo.

Audit AI agent skills, review usage evidence, and compare original, simplified, and baseline workflows.

端到端 QA 验收编排器 · Agent Skill：上下文 → 风险 → 用例 → 脚本 → 执行 → 审查 → 报告，全程自动衔接，只在用例确认时停下等你审核。支持 Claude Code / Codex / Cursor

Open-source AI agents for coding, research and automation. Run locally with your models, persistent memory, multi-agent execution, skills, MCP and self-evolving.

Static analysis for Redis Lua scripts — cluster-unsafe key access, unbounded loops, sandbox violations.

Your AI coding assistant can finally see what you're doing. Local-first screen context for Claude Code, Codex, Gemini CLI and Muse Code.

Codex-style Computer Use for DeepSeek Harness: the official window2 13-method desktop surface plus Chromium automation, a visible overlay with Esc interrupt, per-app approvals and bundled skills. Windows, MIT, no Codex required.

把一张图片逆推成生成提示词（Image to Prompt）：中英双语重建提示词 + 风格标签 + 结构化 JSON prompt，零依赖校验脚本 | Forensic image-to-prompt reconstruction. AGPL-3.0 · Commercial use requires license.

A minimal coding agent harness that fits in the prompt

The control plane for your coding agents, however you run them. One checkout of rules, skills, subagents, hooks and stances, projected into Claude Code and Codex with your own primitives: hook-enforced guardrails, usage telemetry, cost postures and model tiering, under any rules library or orchestration layer.

Find DuckDB query plan anti-patterns — missing predicate pushdown, disk spills, cardinality misestimates.

Full-stack AI short-drama pipeline with FastAPI, Pydantic structured outputs, ComfyUI and FFmpeg

Autonomous, local-first multi-agent AutoML platform. Zero cloud costs, zero API keys, 100% private. Benchmarks 7 algorithms, tunes hyperparameters, and serves live in-browser predictions.

Code Reviewer

The control plane for your coding agents, however you run them. One checkout of rules, skills, subagents, hooks and stances, projected into Claude Code and Codex with your own primitives: hook-enforced guardrails, usage telemetry, cost postures and model tiering, under any rules library or orchestration layer.

⚡️ 一键为 ZCode 注入「自动拉取模型」按钮：批量探测并同步自定义供应商模型列表，样式与原生界面同族。已适配 ZCode 3.12+，客户端升级后自动重新注入

Deterministic repo-wiki build system, driven by any coding agent. Zero LLM, zero network.

Which of your dependencies block a Django upgrade, and in which order to upgrade them.

A production-grade reference implementation for LLM agent systems: flow declared as a LangGraph graph, Pydantic AI for typed agents, generated WebSocket and REST contracts, human approval on irreversible actions, tracing and evals.

tinygrad on Google's Coral Edge TPU: reverse-engineered ISA, our own code generator and USB driver. A whole transformer runs on the stick. Written by AI agents.

Deterministic symbolic codes for embedding vectors. Byte-identical across CPUs and languages, content-addressed, diffable and verifiable. C core with Python, Rust, Go and TypeScript bindings.

Spring Boot + MyBatis 框架感知代码地图：静态分析调用链/SQL/事务/实体影响面，通过 MCP 喂给 AI 编程工具（Trae/Cursor/Claude Code），省 token、防幻觉

Flag ATTACH DATABASE footguns — cross-database atomicity, alias shadowing, ambiguous unqualified tables.

Web map for Satisfactory dedicated servers: live map, production, power, logistics, history and a production planner — self-hosted with Docker in 5 minutes

Baz — Bounded Async Zap. A modern Zig 0.16.0 web framework on bounded/http.

Governed AI agents for sensitive data and business-critical apps. Agents never hold credentials: a gateway checks every call against per-resource permissions, and each answer runs in a sealed container.

AI Terminal Manager — unified session history for Claude Code, Codex, Gemini CLI, opencode and Pi, plus a persistent tmux sidebar

Google Calendar-style weekly view of your Claude Code sessions — local, live, read-only

Self-hosted WhatsApp safety monitor for kids: classifies text, voice notes, images and video with OpenAI Moderation and Whisper, and alerts parents over WhatsApp via OpenWA.

Occam's razor for Claude Code and Codex: a 30-line rule set. Benchmarks vs base: 51-55% lower estimated task cost on Opus/Sonnet 5.5 max with more tests passed. 8.3% fewer tokens in a Sol 6.1 max pilot. Raw data, Docker setup and comparison video included.

3DGS 在线重建查看器 + 个人主页（郑州大学双足实验室 3D 方向考核作品）

Fund-flow pipeline for Robinhood Chain (chain 4663): full ERC-20 Transfer backfill via HyperSync, 1.4B transfers in ClickHouse, read-only JSON API.

Lightweight, async multi-agent framework with native SQLite persistence and handoffs for WhatsApp, Telegram, Webhooks, and CLI. Production-hardened with HMAC webhook security, automatic retries, and turn rollback. Supports OpenAI, Claude, Gemini, and local LLMs.

聚合高校公开招聘信息的 FastAPI + Next.js 日历应用。

Self-hosted market-research terminal: SEC filings, market data on your own vendor keys, and 37 read-only MCP tools for your own AI agent. pip install alphadesk. AGPL-3.0.

AI-powered multi-model workspace and routing system for Claude, Codex and other LLM agents. Provides intelligent prompt routing, agent orchestration, evaluation workflows, CLI integrations, MCP tooling and developer productivity automation.

An enterprise-grade, autonomous AI assistant capable of understanding GitHub repositories, indexing codebases with RAG & vector embeddings, explaining files, detecting security vulnerabilities & bugs, generating unit tests, analyzing issues/PRs, and drafting complete documentation.

Runs the base branch's tests against a PR's code and checks the description's claims against what actually ran. Signed evidence from CI.

研迹 ResearchTrail｜本地 AI 投资研究工作台。串联证券、组合、证据报告、论点复审与 Today 简报，让研究有据、判断留痕。Python 业务内核 · Electron 桌面 · 可复现 Agent 评测。

Скилл для ИИ-агентов и CLI на Python/uv: находит и убирает ИИ-стиль в русских текстах (канцелярит, кальки, штампы), проверяет сохранность правки и оценивает текст по фрагментам в духе «Антиплагиата» с калибровкой по вашим отчётам. Русская адаптация avoid-ai-writing.

Static analysis for Redis Lua scripts — cluster-unsafe key access, unbounded loops, sandbox violations.

True Object-Oriented Python Web Framework.

Local, open-source Jev alternative: browser agent decisions with Laya (System One model) on your own machine. No cloud, no API key. Playwright/CDP, MCP-friendly.

Pemutar & agregator anime mandiri (Web, Windows, Linux, Android). Ringan, tanpa iklan, tanpa akun, dengan ambient stage & subtitle Indonesia.

Write Python. Ship JSON. Run it anywhere, sandboxed.

Coding-agent memory where every fact is a verbatim quote graded by TypeSafe Jev's calibrated confidence. Local-first, SQLite receipts, zero dependencies.

End-to-end apmix.ai account + API key automation. Temp inbox, magic-link signup, automatic Cloudflare Turnstile, and auto-generated API keys — in Python and Node.js.

国科大 Windows 桌面助手：托盘后台运行、轻新课堂、讲座预约、选课规划与自动化、国科大在线视频/PDF；模块化集成与详细使用说明。

Automated MiniMax Agent account signup + daily check-in + JWT/cookie harvester. Uses Zenvex disposable email and Playwright. | Automasi daftar akun MiniMax Agent, check-in harian, dan ambil token/cookies.

Find DuckDB query plan anti-patterns — missing predicate pushdown, disk spills, cardinality misestimates.

从零开始学 AI Agent：17 章渐进教程 + 每章可运行代码 + 算法面试轨（100 题详解）+ 65 道模拟面试题 + 项目实战案例（真实 Go Agent 项目源码级复盘，76 道练习题）+ 本地多语言在线判题（Python/C++）。网页版双击即开，无需 API Key 也能跑完。

🎬 Autonomous AI Studio for Manhwa & Webtoon Recaps — Vision Pipeline (YOLOv8 + Magi v2), Multi-Voice TTS & Video Synthesis

AI-powered multi-model workspace and routing system for Claude, Codex and other LLM agents. Provides intelligent prompt routing, agent orchestration, evaluation workflows, CLI integrations, MCP tooling and developer productivity automation.

Secure LLM gateway and OpenAI-compatible proxy for multi-provider routing, automatic failover, cost optimization, and AI security.

Autonomous AI agent that detects breaking dependency changes via Tree-Sitter AST analysis, synthesizes verified patches, and opens human-reviewed PRs.

这个skill旨在让GPT说话像人，改善防御性回复，不套安慰模板，说大白话。与同类skill对比，说人话的能力排名第一（Codex skill，附盲评对比）

Model routing is a System 1 job: a 421M local decision engine (Laya) against a GPT-5 nano reasoning router. Same accuracy, 35x faster, free.

双击一个文件，给 DeepSeek Harness 注入操作者人格：不拒绝、不中途停、全力执行。只写你自己的文件、官方安装目录零改动、改完下一句对话就生效（不用重启）、改前自动备份可一键还原。纯 Python 标准库，零依赖、零网络请求。

可观测、可评测、可部署的多 Agent 工程平台：结构化路由、Skills、工具/RAG、50 模型实验中心与 LLM-as-Judge。FastAPI + Vue。

Brakes for your AI agents: every tool call is checked against your rules, held for human approval when risky, and recorded in a tamper-evident audit trail. Works with Claude Code and any MCP app.

A focused Playwright helper for setting up one xAI account and creating one API key.

Original self-hosted edition of SEC Watch (archived). The hosted version is at secwatch.app. Website: https://secwatch.app

WorkBuddy 本地配置工具 · rev2.0 — 人格注入 / 云记忆注入 / 命令闸门解锁 / 网页过滤解除。源码交付，纯标准库。 | Local config toolkit for WorkBuddy (rev2.0).

Skill do Claude Code: vídeos de motion graphics feitos em código para qualquer marca (kit de marca, trilha sintetizada, MP4 revisado).

Token-safe AST import re-aliasing and dependency migration engine safely replacing module paths and callsites without regex bugs.

Semantic versioning (SemVer 2.0) breaking change auditor inspecting AST public API surface diffs, parameter reordering, and removed exports.

AST-level static call-graph analyzer and dead code pruner identifying orphaned functions, unreachable methods, and token-bloat clusters.

Monorepo dependency cycle resolver and topological build-order planner using Tarjan SCC and Kahn algorithms.

Exact, read-only model and CLI for Kalshi's Liquidity Incentive Program (built by an autonomous AI agent)

OpenBullet-style bulk cookie & session checker powered by real headless Chromium (Playwright). 139 platform configs, web control panel, CLI, proxy rotation.

High-Performance Mathematical Invariant Firewall & Financial OS for Autonomous AI Agents | Pure Rust, Soroban, ElizaOS & Anthropic MCP

Basic repo with what you need to go from a bare chat interface with your muse agent to an ssh-accessible VM with the nix package manager and full tailscale access (muse ships with a stipped down version),

Self-hosted, single-owner backend for the Cedar read-aloud app (FastAPI + Kokoro TTS). One key, one library, any number of devices.

Tipos e validadores de dados brasileiros de alta performance para Pydantic v2 e FastAPI (CPF, CNPJ 2026, CEP, PIX, Telefone, CNH, PIS, SUS, Título e mais)

Parses unified git diffs, detects security and performance anomalies, generates inline code reviews, and synthesizes unified patches

Parses unified git diffs, detects security and performance anomalies, generates inline code reviews, and synthesizes unified patches

App local e open source para monitorar visitas de gatos e cães à comida, água e caixa de areia com câmeras IP

8 character-driven AI coding agent skills for Codex: debugging, code review, regression testing, and release checks. Weird, but employed. Install: npx skills add SoonGwan/questionable-hires

A platform for running human-centered red teaming evaluations on AI models and systems.

Agent is an local terminal coding agent that can understand coding tasks, inspect and edit workspace files, run commands in a controlled sandbox, ask for approval before risky actions, and verify its work with tests. It uses a streaming LLM provider, a typed event-driven agent loop, safe developer tools, SQLite-backed durable sessions,

ComfyUI 自定义节点：摄影人物提示词随机生成器 · 纯本地 · 不用大模型 · 运行时零联网 · 63 池 949 条中文条目 + 2165 位角色 · 只生成提示词，不生成图片

ComfyUI custom nodes: anime character styling console (clothing / pose / expression / scene slots) with a cross-plugin tag arbiter. Zero third-party dependencies.

MiniMax 反代网关：有头浏览器过滑块批量注册养号 + 账号池自动调度，一键转换为 OpenAI / Anthropic 兼容 API