Best new productivity, design and AI resources: day 27

Twenty-seven days. That's how old Best Productivity Resources is today. Yesterday brought an unusually broad mix of additions, from interface-design references and desktop utilities to AI-agent skills, research workflows and open-source creator tools. Every resource below was reviewed by hand and links straight to its original website or repository.

1. UI Skills, for better interfaces

Want stronger design fundamentals? UI Skills is a focused collection of design-engineering guidance for building better interfaces. Its practical skills cover accessibility, motion, typography, layout, components and overall interface quality, so designers and frontend developers can keep it open as a reference. Link: https://www.ui-skills.com/

2. uncoverLAB, for getting past the blank canvas

Staring at an empty Figma file? uncoverLAB tackles that blank-canvas problem with curated website components, complete app flows and ready-made color palettes for Figma. It's especially useful when you need to get from an early product idea to a structured interface direction quickly. Link: https://uncoverlab.co/

3. Council of High Intelligence, for tough decisions

Tired of getting one agreeable answer? Council of High Intelligence brings structured disagreement to hard decisions inside Claude Code, Codex, Gemini CLI and OpenCode. Distinct analytical personas debate, flag unresolved questions, and produce a verdict with kill criteria and a concrete next action. Link: https://github.com/0xNyk/council-of-high-intelligence

4. FinAI Research, for economics and finance

Doing economic or financial research? FinAI Research is an evidence-first workflow that links literature search, data collection, econometric analysis and verifiable LaTeX drafting. Behind it sits a large collection of data sources, methods, AI skills and journal templates. Link: https://github.com/csmar432/finai-research

5. Obsidian Second Brain, agent memory in your vault

Tired of re-explaining your projects to AI? Obsidian Second Brain gives AI coding agents a durable memory that lives in your Obsidian vault as plain Markdown files. It ships 45 commands for semantic search, self-rewriting notes and scheduled agents that keep the vault tidy. Link: https://github.com/eugeniughelbur/obsidian-second-brain

6. PixelRAG, for visual retrieval

PixelRAG is an open-source take on retrieval-augmented generation for visual information. If you're a developer exploring how AI systems can retrieve and reason over image-heavy documents, and other content that doesn't fit neatly into plain-text search, this one's relevant. Link: https://github.com/StarTrail-org/PixelRAG

7. AutoClip, for turning long videos into clips

AutoClip downloads long videos from YouTube or Bilibili, has an AI model analyse what's in them, and creates highlight clips and compilations. With its web interface and processing workflow, it's a promising self-hosted option for creators who regularly repurpose long-form video. Link: https://github.com/zhouxiaoka/autoclip

8. DeepTeam, for red-teaming AI

DeepTeam is an open-source red-teaming framework for testing AI systems. Teams can use it to probe models and agents for safety, security and reliability weaknesses before real users ever get near them. Link: https://github.com/confident-ai/deepteam

9. pdfcn, for PDFs from React

Need to generate PDFs from a React app? pdfcn is a free library of PDF components built on Takumi and Forme, with a shadcn-style workflow and one-command setup. Link: https://github.com/shadcn-labs/pdfcn

10. Ghostty, for a faster terminal

Spend much of your day in the terminal? Ghostty is a fast, native terminal emulator designed around platform conventions and a modern rendering stack. It's a strong pick if you still want a polished desktop experience. Link: https://github.com/ghostty-org/ghostty

More good additions from yesterday

There's more in the directory too. shadercn brings reusable WebGPU shader components. Awesome GNOME helps you find Linux desktop apps and extensions. MonitorControl and Loop cover practical Mac workflows. Awesome Academic Research Skills is for AI-assisted scholarship, Builder.io Skills and Agent Reach are for agent workflows, and Story Skills supports structured storytelling work.

So, where should you start?

With whatever problem is already sitting in front of you. Doing interface work? Try UI Skills or uncoverLAB. For agent memory in your notes, Obsidian Second Brain. For repurposing video, AutoClip. For rigorous research, FinAI Research. For a faster daily development setup, Ghostty. Browse every addition at https://bestproductivityresources.com/directory