hupicake
-
Built with Claude pixelfx: 244 code-drawn pixel art effects and a Claude Code skill for making more
pixelfx (patch-gg/pixelfx) is an MIT-licensed library of 244 pixel art effects written as small JavaScript files, with sprite sheet export for any engine and a Claude Code skill that drafts, renders and refines new ones.
-
Built with Claude Pinky lets AI agents use your iPhone or Android phone, with approvals and blocked apps
Pinky is an open-source desktop app that lets Claude Code and other MCP agents operate a real iPhone or Android phone while you approve risky taps, block banking apps and keep recordings on your computer.
-
News Can individuals get Claude Mythos 5.2? What the Cyber Verification Program allows
Individual security researchers can apply for Defense Access in Anthropic's Cyber Verification Program, and Anthropic's help article lists Claude Mythos 5.2 among its models. Here is who qualifies, what the application asks for, the December 15 security deadline and what Mythos costs on each plan.
-
Built with Claude Free My TV: an open-source, ad-free home screen for Google TV, LG and Samsung, built with Claude Code
Free My TV (simbasjungles/free-my-tv) replaces the stock Google TV and Android TV home screen with an ad-free one and runs as an app on LG and Samsung TVs. It is MIT-licensed, has about 6,400 GitHub stars, one button to go back, and most commits carry a Claude Co-Authored-By line.
-
News Fable 5.2's free week on Claude Pro ends November 16; usage credits after that
Claude Pro and standard Team seats can use Fable 5.2 at no extra cost until November 16 at 11:59 pm PT, up to half the weekly limit. After that, Fable usage on Pro is billed to usage credits, and the launch-week Fable reset expires at the same moment.
-
Daily DeepSeek opens its app features as API tools; study finds Lovable apps leaking user data
DeepSeek now sells the search, file, voice and safety features behind its chat app as API tools, with prices you can plan around. And a scan of 159,041 Lovable projects shows how often Supabase-backed apps let strangers read private records, plus how to check yours.
-
News DeepSeek opens its app's search, file, voice and safety features to the API
DeepSeek now exposes the features behind its chat app through its API: server-side web search, URL reading, file parsing and safety checks on deepseek-flash, plus standalone Search, Fetch, OCR, ASR and TTS endpoints.
-
News Study: thousands of Lovable apps left user data readable through Supabase
A Crissy Field study of 159,041 Lovable projects found 7,403 sites where Supabase tables holding email addresses could be read without logging in, including passport details and 200,000 chat messages. Here is how to check your own app.
-
Daily Claude Code mods reach the desktop app's colors, and GitHub already counts 2,300 of them
A mods special: the desktop app's Code tab can now be recolored by a mod, community mods pass 2,300 on GitHub, and a close read of one bundle's 78% usage claim.
-
News Claude Code mods can now recolor the desktop app's Code tab
A Claude Code mod can now set the colors of the Claude desktop app's Code tab, on Claude Code 2.1.324 or later. Here is how to get one, what stays the same in the terminal, and how to take the colors off again.
-
Tools Six weeks after launch, about 2,300 GitHub repos call themselves a Claude Code mod
Six weeks after Claude Code mods launched, about 2,300 GitHub repositories describe themselves as a Claude Code mod, and hundreds of them mention usage. The usage meters read your limits in three different ways.
-
Research ystack says its Claude Code mods cut usage 78%. Its own write-up calls that the upper end
ystack, a new bundle of 35 Claude Code mods, says it cut usage about 78% per hour of work. The figure comes from one heavy user's logs, and the author treats 39% per API call as the firm number.
-
Tips What a Claude Code mod can do on your machine, and the guard many solo users don't get
Claude Code mods run unsandboxed, with your permissions. A built-in guard stops them overriding your deny rules, but by default it only loads with managed settings, a managed-mcp.json file, or a Team or Enterprise login.
-
Tips How to have Claude Code write a mod for you, and keep it after the session ends
You can describe a Claude Code mod in one sentence and have Claude write it. The catch is that it lives in a per-session folder that gets cleaned up. Here is the whole loop, and how to keep the mod.
-
News Self-hosted Claude Managed Agents: what you run, and what stops working
Anthropic opened a private beta of self-hosted Claude Managed Agents on November 12: the server, database and sandboxes run in your own cloud or data center. Its docs also list the hosted features you lose.
-
Daily Google's Gemini 4 Hydrogen gets cheap only after December 1, and Claude Code tests a Code Mode
Gemini 4 Hydrogen brings Gemini 4 to every API tier at $1.50/$7.50, and Orchestrator pairs it with Carbon in Antigravity. Claude Code 2.1.323 adds an experimental Code Mode. Also: Edge0's 145B local model, a GPU fingerprint study of Chinese models, and Brave's pay-per-search API.
-
News Gemini 4 Hydrogen costs $1.50/$7.50, and it only gets much cheaper than Carbon on December 1
Google released Gemini 4 Hydrogen, a fast Gemini 4 model at $1.50 per million input tokens and $7.50 output, on every Gemini API tier including free. Today it is 25% cheaper per token than Carbon; from December 1 the gap grows to 62.5%.
-
Tools Claude Code 2.1.323 adds an experimental Code Mode and version ranges for plugins
Claude Code 2.1.323 adds Code Mode, an opt-in experiment where Claude writes JavaScript that makes several tool calls in one step, and an engines field that lets plugins declare which Claude Code versions they support. The Code Mode docs are now live: here is how to turn it on and its limits.
-
News Edge0-145B: pitched for laptops, but no laptop appears in its speed table
Edge0 released Edge0-145B and Edge0-120B, sparse models derived from Qwen3.8-Flash-Next that keep hot experts on the GPU and stream the rest from RAM and disk. The launch post talks about laptops; the speed tables list RTX cards, Spark and an M2 Ultra. Setup, sizes, licence and caveats.
-
Research Researcher says released weights show six Chinese models were started on Nvidia Blackwell GPUs
Will DePue's checkpoint-forensics report says the starting values left in published weights reveal which GPU created them, and that six models from DeepSeek, Moonshot, MiniMax and Ant Group match Nvidia Blackwell. How the method works, and what it cannot show.

