← Feed

Posts

• Post #2686977

  Figjam MCP is something I have been needing in my life for a very long time. From mind maps, to workflow process maps, to opportunity solution trees, I am always in Figjam nudging around nodes and connectors. https://www.figma.com/release-notes/?title=figjam-diagramming-in-claude

• Post #2686976

  After two days of Moltbot (Clawdbot), I am starting to feel it's not for me. At least not yet. I'm struggling with having Moltbot achieve things on particular days/times, plan instead of execute, and most of all to stop pretending to be me all the time. Will keep trying for now.

• Post #2686975

  Figma just released user groups for easier sharing and permissions. As a freelance designer, this will make working with different clients so much easier. https://www.figma.com/release-notes/?title=user-groups-now-available

• Post #2686974

  Homelabbing finally feels like a superpower in 2026. Stack: OpenClaw, Ollama, Open WebUI, n8n, Tailscale, OpenCode, Penpot, & Obsidian. Model of the moment: minimax-m2.5. What's in your homelab stack?

• Post #2686973

  What if your AI wasn't just for work? Just saw Jesse Genet on How I AI using Obsidian, OpenClaw, and a squad of Macs for homeschooling, finances, and total life management. It's making me rethink patterns in my own life such as home inventory management. https://www.youtube.com/watch?v=96Vl8s3EQhk

• Post #2686972

  Say hello to Homie. Using n8n to create AI Agents has been so much fun. I'm currently working on a publishing assistant for my homelab newsletter project. I'll keep you guys updated as I progress.

• Post #2686971

  I found this interesting. Google has implemented the ability to download Google Docs as Markdown files. With AI becoming a growing presence in our lives, and with Markdown as its primary language, it's nice to have tooling that makes the transition easier.

• Post #2686970

  Brad Frost is always a great follow. His Real-Time UI post demos agent-to-UI live during a meeting, a workflow that could collapse idea-to-execution from weeks to days and kill misalignment before it starts. Check it out: https://bradfrost.com/blog/post/real-time-ui/

• Post #2686969

  I've been waiting for Figma Slots for years! No more creating placeholder components with complicated visibility Boolean properties for my design systems. This is so huge for design workflows. https://youtu.be/fpc0cUt1y1M

• Post #2686968

  Pencil.dev has been one of the most interesting AI-powered design tools I've come across. The latest parallel agent functionality seems particularly powerful. I'd love to see a deeper integration with Ollama/OpenCode, however. Check out the walkthrough. https://www.youtube.com/watch?v=w4RY7PnfRU8

• Post #2686967

  OpenUI is an interesting approach to Generative UI. It focuses on providing a token-efficient, structured markup language for UI components and layouts. The demo is fun but I'm itching to throw some custom components at it. https://www.openui.com/

• Post #2686966

  I've been thinking a lot about the new product designer workflow in the age of AI. Agentation seems like a sneak peek into that future. Identify a design change in live dev, comment on the exact element you need changed, have AI fix it in the background. https://www.agentation.com/

• Post #2686965

  Zed is redefining accessible type design. Text and display variants, 435 languages, Braille support, and lab-tested legibility for low-vision readers. This is the standard all type systems should be held to. Inclusive by design, not as an afterthought. https://www.typotheque.com/blog/zed-a-sans-for-the-needs-of-21century

• Post #2686964

  Designers tokenize color, spacing, and typography but what about motion? Still rare. Francesco Improta shares a tiered approach, including accessibility considerations, that seeks to solve this overlooked part of design systems. Really a good foundation. https://designtokens.substack.com/p/motion-tokens-naming-your-movement

• Post #2686963

  No mouse November, anyone? Every now and then my wireless mouse dies on me and I am forced to learn how bad the web is without the use of a mouse once again. I'm all for raising awareness for those who experience the web this way on the daily. https://nomouse.org/

• Post #2686962

  The Design Systems Report for 2026 by zeroheight is out! Loads of fantastic statistics for fellow designers to understand the current state of design and design systems. One stat that hit me was how under-represented accessibility and PM are. https://report.zeroheight.com/

• Post #2686961

  Figma has rebuilt component instances using Materializer. Large design systems are now 50% faster. Separating layout and variable logic unlocks features like slots and rich text. This is a huge win for design ops scalability. https://www.figma.com/blog/how-we-rebuilt-the-foundations-of-component-instances/

• Post #2686960

  Claude is releasing features like no other lately. The latest is a feature that allows Claude to completely take over a computer. While I won't be doing this on my actual machine, I am very tempted to try it on a virtual machine. https://x.com/claudeai/status/2036195789601374705

• Post #2686959

  If you have been wondering what exactly LLM quantization is or why TurboQuant is starting to become all the rage, I have never seen it described in a simpler or more approachable way than this. Learn yourselves! https://www.reddit.com/r/LocalLLaMA/comments/1s62g5v/a_simple_explanation_of_the_key_idea_behind/

• Post #2686958

  The OpenAI image generation models prompting guide is a great resource for really any image generation model. A lot of great techniques and examples are included. Check it out. https://developers.openai.com/cookbook/examples/multimodal/image-gen-models-prompting-guide