hupicake
-
Research FrontierHarness Eval v1.1: with open models, Claude Code was among the costliest harnesses per solved task
Runta and Baseten ran 22 configurations of 18 coding-agent harnesses with five open models over 62 terminal tasks. The best-value harness changed with the model, and Claude Code was the second most expensive per solved task. Closed models were not tested.
-
Tools ai-manju-skills: 21 AI comic-drama templates that turn one idea into a full series plan
ai-manju-skills is a set of 21 Chinese-language AI comic-drama templates that turn a single story idea into a production plan, asset prompts, episode scripts, storyboard and video prompts, and a release pack.
-
Research Claude Code on GitHub by the numbers: 121,725 repos, 24,336 of them new in 30 days (November 2026)
121,725 public GitHub repositories carry the claude-code topic as of November 17, 2026, and 24,336 were created in the last 30 days. Counts by category, stars, language and month, the top repos, the week's risers and the queries behind the counts.
-
Tools BeautifulRust: Agent Skills for writing, reviewing, and porting idiomatic Rust
BeautifulRust is a set of Agent Skills that teach coding agents to write, review, and port idiomatic Rust, with a tested clippy config to enforce the rules.
-
Tools cinematic-skill: turn a script and a voice into a narrated cinematic film built from code
Moh4696/cinematic-skill is an agent skill that turns a script and a voice into a narrated cinematic film built entirely from code, with one continuous camera over photos, maps and cards.
-
News Hugging Face Inference Providers now route through OpenRouter: what changes for your code, keys and Claude Code
Hugging Face now sends language model requests from Inference Providers through OpenRouter's router, with automatic retries on another provider. Your code and token stay the same, but your own provider API keys no longer work for language models.
-
News macOS 27.3 beta splits Device Control and Data Access and lets you scope Full Disk Access: what it means for AI agents
The macOS 27.3 beta splits Device Control and Data Access into System & App Protection Override and Accessibility. For some apps you can now limit Full Disk Access to chosen folders and app control to chosen apps. Existing grants stay in place.
-
News Claude Meetings beta: how Claude joins Google Meet, uses your connectors and what it costs
Claude Meetings puts Claude into Google Meet calls on Team and Enterprise plans to take notes, answer questions, build docs and use your connected tools with permission. Here are the settings, limits and what counts toward usage.
-
News Anthropic moves to 30-day bug disclosure for high-severity reports, three days after Project Zero dropped 90+30
For vulnerabilities it reports on or after January 1, 2027, Anthropic will disclose high- and critical-severity bugs after 30 days and all others after 45, down from 90. Google Project Zero has moved to a fixed 90-day clock that steps down to 60 and then 30 days.
-
Daily Anthropic moves to 30-day bug disclosure, and Google quietly patches Gemma 4 tokenizers
Anthropic will publish high- and critical-severity bugs 30 days after reporting them from January 2027, a week after Project Zero dropped its 90+30 policy. Google also fixed glitch tokens in the small Gemma 4 tokenizers. Plus today's fastest-rising open-source projects for Claude Code users.
-
Tools Google patched glitch tokens in the Gemma 4 E2B and E4B tokenizers: how to check your copy
Google patched tokenizer.json for Gemma 4 E2B and E4B on November 3, 2026, making 217 to 233 glitch tokens unreachable so they stop garbling output. Weights are unchanged, and the 26B and 31B repos have not been patched yet.
-
Tools open-code-review: a Go CLI that reviews Git diffs with a model and line-level comments
alibaba/open-code-review is a command-line code review tool that reads a Git diff, sends the changed files to a model you choose, and returns comments tied to specific lines.
-
Tools Plannotator: review your coding agent's plans, diffs and messages before it acts
Plannotator is a local, browser-based review workspace for AI coding agents that lets you annotate plans, diffs and agent messages and send the feedback back to the agent.
-
Tools Univer: embed a spreadsheet and document editor in your own app
dream-num/univer is a TypeScript SDK that embeds a spreadsheet and document editor in your own web product, with the same content models for browser editing, headless Node.js work, and agent-driven edits.
-
News New BIS rule targets remote access to AI chips and adds license requirements for Singapore, Malaysia and three more countries
BIS's November 16, 2026 interim final rule requires a license when advanced AI chips knowingly feed data centers giving Macau- or D:5-headquartered firms over 12,000,000 aggregate TPP, and for exports to Singapore, Malaysia, Indonesia, Thailand and Brazil.
-
Tools AG-UI: connect an AI agent to a frontend app with a shared event protocol
AG-UI is an event protocol that connects AI agents to user-facing applications. This post covers the scaffold, the packages, the key the examples need, and the version 1.0 limits to watch.
-
News Qwen4-Max is out: 1M context, $1.50/$4.50 pricing, and a Claude Code setup
Alibaba released Qwen4-Max on November 16 at $1.50 in and $4.50 out per million tokens, 25% below Qwen3.8-Max (0902), with cache reads 70% cheaper. It has a 1M-token context and works in Claude Code through QwenCloud's Anthropic-compatible endpoint.
-
Tips Anthropic's always-on agent guide: the five parts, and a minimal build plan
Anthropic's November 13 guide says an always-on agent is mostly asleep: short runs started by inbox events, including timer events, with memory, permissions and an outbox around the same agent loop. Here are the five parts and one way to build them.
-
Tools Kanzo puts a kanban board and an MCP server for coding agents in one 4.7 MB binary
Kanzo (emanuele-em/kanzo) is a self-hosted kanban board and MCP server in one 4.7 MB Rust binary, so you and coding agents like Claude Code and Codex take, move and comment on the same tasks, on one computer or a shared board. AGPL-3.0.
-
Built with Claude TV Detox gives an old Android TV a 30 MB ad-free home screen and a Claude Code cleanup you can undo
TV Detox (Ansub/tv-detox) is a free, MIT-licensed kit for slow Android and Google TVs: a tiny ad-free home screen app plus a prompt that has Claude Code or Codex turn off preinstalled apps over ADB, asking before each change. Here is how to install it, the risks and how to undo it.

