2026年10月1日 技术热点总结

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

技术栈：Python、ai-agent、ai-agents、ai-browser、anti-detect-browser、browser-agent、browser-automation、chatgpt、dotfiles

核心介绍：dots Every AI agent is a model and a browser. You can swap the model with one flag. The browser is what the website sees. Windows, in PowerShell: powershell -ExecutionPolicy ByPass -c “irm https://astral.sh/uv/install.ps1 | iex” $env:Path = “$env:USERPROFILE.localbin;$env:Path” uvx –from git+https://github.com/feder-cr/dots dots –openrouter-key sk-or-… curl -…

项目简介：A tiny friend that lives in your notch (macOS) or at the top of your screen (Windows) and keeps an eye on your Claude Code sessions.

技术栈：Swift、ai-agents、anthropic、claude、claude-code、dynamic-island、macos、macos-app、menubar-app

核心介绍：Approve permissions, watch your agents work, drop a file, chat with Claude — all without leaving what you’re doing. Some studios showed off gorgeous notch companions… and never let anyone use them. Meet Mochi: a soft little squircle with big eyes that pops out of your notch, waves hello, follows your cursor with its eyes, gets annoyed when y…

项目简介：A comprehensive logo-design skill for Claude, Gemini CLI, Codex and other AI agents: principles, process, SVG craft, testing tools and a 1,400+ logo reference library.

技术栈：HTML、agent-skills、branding、claude、claude-skills、codex、gemini-cli、logo-design、svg

核心介绍：A comprehensive logo-design skill that turns Claude — or any agent that supports Agent Skills, such as Gemini CLI, Codex CLI, Cursor or GitHub Copilot — into a disciplined identity designer, from the first brief to production-ready SVG files and brand guidelines. geometric construction, optical corrections (oversho…

项目简介：A growing collection of viral videos made with Claude Opus 5.5 and the prompts behind them. Watch each original next to a live remake on Skillry. Updated regularly.

技术栈：ai-video、awesome、awesome-list、claude、claude-opus、creative-coding、motion-graphics、prompts

核心介绍：Viral videos people made with Claude Opus 5.5, each with the prompt the creator shared (or, where no prompt was published, their original post). Every video here was made by asking Claude Opus 5.5 to write the animation as code (HTML, Canvas, SVG, Three.js and friends). Pick one, copy its prompt, and try it wi…

技术栈：HTML

核心介绍：纯前端的户型装修设计工具：在 2D 平面图上摆放家具、拆改墙体、测量尺寸，一键切换到 Three.js 3D 场景，可以鸟瞰，也可以第一人称漫游。整个应用就是一个 index.html，无需构建，打开即用。 git clone 然后直接用浏览器打开 index.html。也可以起一个本地静态服务器： python3 -m http.server 8000 > Three.js 通过 jsDelivr CDN 加载，首次打开 3D 场景需要联网。 户型数据写在 index.html 里： 改这些数据就能换成自己的户型。

关键特性：按原始户型 1:60 / 1:100 比例显示，尺寸单位 mm；从左侧家具库拖入 60 余种家具家电（卧室、客厅、餐厨、卫浴、家电、书房休闲）；拖动移动、旋转（Shift 自由角度）、调整尺寸，贴墙自动吸附；测量工具（靠近墙面自动吸附，Shift 锁定水平 / 垂直）；拆改非承重墙，承重墙单独标示。

项目数据：⭐ 937 Stars，🍴 207 Forks

项目简介：Point Claude at any game. Skills, tools and the fal MCP that let Claude Code mod almost any PC game you own: recon, reverse engineering, fal-generated art/3D/audio, in-game testing, showcase videos.

技术栈：Python、age-of-empires、claude-code、claude-code-plugin、fal、game-assets、game-modding、mcp、modding

核心介绍：Skills, tools and a shared knowledge base that let any AI coding agent mod almost any PC game you own. Works with Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, OpenCode, or anything that reads AGENTS.md. The agent finds the game, works out the engine and the route, …

Browser-use agents often carry information in their context as they move between websites. While it may be necessary for task completion, it also creates a privacy risk, especially when the information contains a private fact regarding the user. For example, an agent may learn a user's affiliation after reading a membership record. If it later selects a registration option specific to that affiliation on another website instead of a general option, the information gets leaked. In this work, we define and study behavioural side-channel leakage in browser-use agents, where an agent's actions …

Diffusion models can produce striking images and videos, but they still struggle with the compositional details that make a generation faithful to a prompt, such as object counts, attribute binding, spatial relations, and temporally grounded actions. A common way to improve prompt satisfaction is to spend more compute at test time through Best-of-N sampling, but final-sample selection is fixed. Best-of-N can only choose among completed outputs and cannot repair a promising trajectory before it fails. We introduce PreviewDiff, a training-free test-time search method that turns diffusion samp…

Looped language models (LoopLMs) increase computational depth through parameter sharing, offering a path to scale inference computation without adding parameters. However, it remains unclear when additional recurrence is beneficial and how architectural choices affect its effectiveness. Through controlled experiments, we systematically examine (1) when recurrence helps, (2) where it should be applied, and (3) how its conditioning affects performance. Our evaluation covers inference budgets below, within, and beyond the training horizon under knowledge and reasoning tasks. (1) We find that r…

State Space Models (SSMs) compress sequence history into a bounded recurrent state, making the resulting memory law a central architectural choice for long-context performance. Most modern SSMs rely on ODE-based dynamics that lead to exponential forgetting, limiting their ability to retain information over broad temporal ranges. We introduce FRAC, a selective SSM architecture derived from fractional dynamics that replaces this exponential decay with power-law long memory. To make fractional dynamics practical, FRAC approximates the heavy-tailed target kernel with a finite-state, log-spaced …

Chunked KV-cache compression reduces the memory and attention costs of long-context inference by compressing windows of consecutive tokens into fewer cache entries at a fixed stride. Such compression also introduces a new positional coordinate: a token's phase, or its position relative to compression-window boundaries. We uncover a systematic asymmetry in models using such compression: the same information can be easy to retrieve at one phase and difficult at another. We call this periodic variation in retrieval performance phase sensitivity. In large open-weight models with such compressio…

Language model agents serving organizations must coordinate requests from multiple users while using knowledge distributed across their interactions. We identify two complementary capabilities for this setting, namely cross-user interaction and decision-making, as well as cross-user memory and knowledge use. Both capabilities are governed by organizational constraints across three aspects: user identity, authority, and access permissions; the attribution and temporal validity of information; and rules for resolving conflicting requirements across users and completion requirements for joint …