gerdian
28 posts

gerdian
@bvnine
sht posting - dont take it seriously

Ok, but can you write a hook for your coding agent harness of choice, that let's you play snake directly in the tui, while the agent works? With pause/resume? With game state saved to session file? Thought so. Available soon in pi github.com/badlogic/pi-mo…

Z.ai is set for its IPO on Jan 8, 2026. This journey has been powered by our developers, researchers, and users from Day 1. Thank you for building this reality with us!




/review in Droid You asked, we delivered. Review a branch, commit, or uncommitted changes. Add custom instructions to focus on what matters most to you. Droid analyzes your changes, identifies bugs and improvements, and ranks each finding by priority (P0-P3) from 'drop everything' to 'nice to have' - so you know exactly what to fix first. Try it out and let us know what you think!

We've doubled the Vibe context limit from 100k to 200k. Happy shipping! → uv tool install mistral-vibe





The real q: is @openclaw my assistant, or am I their assistant.

Hey Charmers! Crush has debuted another killer release 💅 Here’s what’s new: ✨ Crush can now run tool calls in parallel (pretty glamorous, huh?) We’re ALL about saving you some time 😊 Check it out below! ✨ We’ve optimized how Crush runs with @zai_org’s GLM models! GLM-4.6 is SUPER popular amongst our users and we wanted to make that experience even smoother! 💖 ✨ We’ve tuned how system prompting operates for all models across the board 🎀💘 Let us know how you like it! That sums it up, but if you want to read more 📖 click here: github.com/charmbracelet/…

Introducing the Mistral 3 family of models: Frontier intelligence at all sizes. Apache 2.0. Details in 🧵

Anthropic is acquiring @bunjavascript to further accelerate Claude Code’s growth. We're delighted that Bun—which has dramatically improved the JavaScript and TypeScript developer experience—is joining us to make Claude Code even better. Read more: anthropic.com/news/anthropic…



🤖 Thinking in Tool-Use 🔹 Introduces a new massive agent training data synthesis method covering 1,800+ environments & 85k+ complex instructions. 🔹 DeepSeek-V3.2 is our first model to integrate thinking directly into tool-use, and also supports tool-use in both thinking and non-thinking modes. 3/n







