Five things worth your attention if you build with AI, about 3 minutes. Join free and get 13 ready-to-paste Starter Prompts for coding agents, vibe coding and game dev right away.
MODEL
1 Tencent ships Hy4 with vision and computer use
Hy4 is the first Hy model that can see, and the first that does computer use across CLIs and GUIs, with targeted training on tools like Blender and CAD. It's live on Tencent Cloud and OpenRouter at $0.834 input and $2.501 output per million tokens, and Tencent says it plans to release the weights later with a system card.
Why builders care: A cheap, soon-to-be-open model aimed at agents and game dev. Tencent itself says a notable gap to top proprietary models remains, so test before you switch.
2 GitBook says an attacker used AI agents to get in
An attacker using AI agents got some of GitBook's infrastructure credentials on Nov 2. Integration tokens, Git Sync secrets, API tokens and Assistant visitor questions may have been accessed. GitBook revoked all personal API tokens and restored service Nov 3.
Why builders care: If your docs live on GitBook, rotate your GitBook and third-party tokens and re-authenticate its MCP server today.
After deleting "a couple terabytes of ai slop," itch.io is adding generative AI disclosure filtering to more of the site, including search results. Tabletop games, books, comics and soundtracks now need an AI disclosure before you can create or edit a page, joining game assets.
Why builders care: If you sell on itch.io, fill in your disclosure accurately. It now decides who finds you.
4 Crouton Alpha: a free stealth model on OpenRouter
OpenRouter listed Crouton Alpha on Nov 3. It's free, with a 1,048,576-token context, multimodal input, tool calling and structured outputs. The provider is anonymous, and prompts may be retained, though not used for training.
Why builders care: A free long-context model to try for agent sub-tasks. Just keep secrets and client code out of it.
5 QuickProof: write C specs like tests, let agents prove them
GenProof's QuickProof turns a small C program into a Rocq theorem that AI agents then prove, and qp verify checks that nobody weakened the statement. GenProof says it found more than 300 bugs across 20 C projects, including Redis and zstd.
Why builders care: Practical formal verification for C. Note that it's under the PolyForm Noncommercial license.
That's the issue. Know a builder who'd use this? Share it on X or forward them this page. New here? Subscribe free for tomorrow's.
Get the next issue in your inbox
Five things worth your attention in AI, agents, vibe coding and game dev, every weekday. Plus 13 free Starter Prompts and the AI Builder Starter Stack PDF when you join.
