A curated list and taxonomy of test-time robot learning: RL post-training, policy steering, test-time adaptation and training, test-time scaling, in-context learning, and policy self-improvement.
HBG:把 Gemini Web + Flow/Veo 图片、视频能力统一封装进 Docker 与 CLI,支持本地浏览器授权、自动下载与可选媒体后处理。
—
—
Read and reset Epson waste ink counters. Free and open source. Does not empty your pads - it only resets the counter.
Curated MCP server for Nextcloud: files, calendar, notes, deck and contacts as read-first tools. Never deletes, moves or re-shares; edits only opt-in and ETag-protected.
Corrects your shell command before it runs, or after it fails. Explains itself, learns from you, works with coding agents. Successor to The Fuck.
AI watermark removal built for Apple Silicon. SynthID-class pixel regeneration, visible vendor labels and provenance metadata, with every pipeline decision measured on Metal.
—
Servidor MCP oficial do dados.ons.org.br . Permite a assistentes de IA buscar, explorar e analisar os datasets do Portal de Dados Abertos do ONS, o Operador Nacional do Sistema Elétrico.
Translation tooling for energy-system model formats
Ferramenta robusta de monitoramento, análise, segurança e inteligência de fontes abertas (OSINT).
谁说AI不能生成GIF? 本skill利用GPT-Image 2.5 一次性生成多格帧图 - 合成为动画gif - 一次调用就生成动态图。产品动图生成器 > 你给我产品照片,我还你一段会自动循环播放的产品动图。 这个技能是给"不会用设计软件、但需要动图素材"的人用的。 你不需要懂任何技术,只要会发图片、会说话就够了。
A tiny jev-like model that answers Choice, Score and Noul questions in one forward pass and returns calibrated probabilities. MLX or PyTorch, fully offline, System One compatible.
Ultimate assistant for ARC Raiders. Features: Infinite Health, Unlimited Stamina, Fast Resource Collection, Instant Crafting, No Hunger/Thirst, Unlimited Ammo,One-Hit Kill, Damage Multiplier, Infinite Currency, and more. 15+ functions.
MCP-сервер правил обмена 1С через агента: «Конвертация данных» 2 и 3 (EnterpriseData) — разбор, сопоставление, правка и проверка правил
吉星派对 个人和对局信息查看工具
AI API gateway / relay: 307 models behind one OpenAI-compatible key. image2.5 $0.0085/image, Seedance 2.0 Mini $0.01056/sec, LLM from $0.0228/M tokens. USD settlement, pay-as-you-go, $1 minimum top-up.
—
Repositorio oficial de skills para el entorno Mexicano caótico.
Конвертер профилей zapret-discord-youtube для nfqws2 на Keenetic и OpenWrt
—
—
A long-form roleplay engine for one human and one AI. Slot-machine world generation, summon-based narrator & NPCs, and layered redlines you can edit without touching code.给一个人和一个 AI 的长线角色扮演引擎。摇老虎机开世界,旁白和 NPC 召唤制出场,红线分层——改配置文件就能换玩法,不用碰代码。
Vanilla RAG baseline for the Agent Memory Leaderboard
—
CUMCM 数学建模 AI Skill|支持 Codex、Claude Code|深度读题、EDA 探索性分析、逐问建模、方案评分、文献与数据核验、Python 求解验证、优化/预测/仿真与完整竞赛交付
100+ AI, Gen AI, LLM, Agentic AI, Forward Deployed Engineer (FDE), AI Systems Architect, Applied AI, LLMOps, AI Platform engineering interview questions across 10 topics - short answer, detail, worked example, and interview tips for each.
Local desktop app to build structured MiniMax H3 video prompts (T2VA/I2VA/FL2VA/L2VA/Ref2VA) using a local LLM via Ollama or LM Studio. Streams output, auto-describes reference images with vision models, 100% offline — no API keys, no cloud.
The same HTTP API in 10 languages at 3 sizes, 30 implementations. Token cost for LLM agents, plus cold start, latency, throughput, CPU and memory. All measured, all verified over live HTTP.
Local Ecobee control in Python via HomeKit. No cloud, no ecobee API key, no Home Assistant. Pair once, then read/control temp, mode, sensors, and more from a GUI, CLI, or library.
Malva algorithm for indexing and index querying
—
WiFi CSI human sensing for Python presence, fall detection, breathing rate, activity classification, and occupancy counting from an ESP32. No camera, no cloud, no PyTorch/CUDA at runtime.
—
—
—
The agent that evolves with you 𖤍
WioSpor CloudStream live TV repository
⚡️ 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
—
ChangMing Serif TC: a humanist serif webfont for Traditional Chinese, built on Noto Serif TC with softened serifs, rounded turns and lower stroke contrast. Covers Taiwanese Hokkien and Hakka characters and romanization. Sliced unicode-range web kit, TTF/WOFF2 and a WordPress plugin. 昌明體:以思源明體為骨架的繁體中文人文明體,涵蓋台語、客語用字,提供分片網頁字型、TTF/WOFF2 與 WordPress 外掛。
—
一个本地可扩展个人 AI agent 工作台(a local-first, extensible personal AI agent workspace)
—
Installable Codex skill for training MiniMax H3 Ref2VA LoRAs on Google Colab
[NIPS 2026🔥] DC-SAE: Deep Compression Semantic Autoencoder for Faster Diffusion Convergence
WorldGuide: Goal-Directed Video World Model for Procedural Task Execution.
Ultimate assistant for Once Human. Features: Unlimited Health, Unlimited Stamina, XP Multiplier, All Resources, Blueprints, No Crafting Requirements, 15+ functions.
LLM engine with adaptive expert loading. DeepSeek V4.1 Flash (510 GB) on two DGX Sparks at its released MXFP4 weights.Or GLM 5.3 full EXL3.
Official code for FoveaVLA: Foveation Actions Replace the Wrist Camera in VLAs
—
iCloud Hidden Open Menu Bypass No jailbreak. MobileGestalt exploit for iOS 15.0–18.7.4 & 26.2 Beta 1. Hides Apple ID, removes password prompt. Does NOT disable FMI. Requirements: Open Menu, Wi-Fi. Steps: run tool → reboot → Reset All Settings (NOT Erase!) → sign in new ID. ⚠️ Don't update to 26.2 Beta 2+ (patched). Legal use only.
Embedded sensor telemetry with signal quality checks, buffering and cloud adapter boundaries.
This project is simply a demonstration of how to obtain the Pair Extraordinaire Badge in an authentic, genuine, and up-to-date way
Save GitHub repos you scroll past into your own private repo; analyzed by your own Claude subscription
Zero-touch session rotation and supervision for Claude Code
Research software for reconstructing 3x3 puzzle-cube solves from 120 fps phone video, with a local workbench, dataset, and models.
An open-source interactive audio workstation that learns the harmonic structure of a sample library. Reflector absorbs your sound world and makes it navigable across sessions and individual arrangements.
—