AI Skills

Coach for open source project decisions using official Open Source Guides as a framework.

Translate repository READMEs into multiple languages while preserving structure, links, and code.

Explore user intent and design before implementation—mandatory first step for any creative work.

Access the user's live browser session for tasks requiring rendered state, interaction, or authentication.

Compact conversations into handoff documents for seamless agent-to-agent transitions.

Assess Kubernetes workloads and cluster configuration for AKS Automatic compatibility.

Accelerate downloads and package installs by rewriting URLs to your Xget proxy deployment.

CLI tool for creating, extracting, and managing .tzst and .tar.zst archives with compression, filtering, and conflict resolution.

Upload and download encrypted files via Xdrop server from the terminal.

Render MP4 videos from React/Remotion component code via inference.sh CLI.

Build throwaway prototypes to validate design decisions and state models quickly.

Create AI avatar and talking head videos via inference.sh CLI with built-in TTS and lipsync.

Generate AI videos with 40+ models (Veo, Seedance, HappyHorse, Wan, Grok) via inference.sh CLI.

Generate images with 50+ AI models including FLUX, GPT-Image-2, Gemini, and Grok via inference.sh CLI.

Automate Twitter/X posting, engagement, and user management via inference.sh CLI

Create, edit, and manage Feishu slides with full page control and visual design support.

Comprehensive UI/UX design intelligence with 50+ styles, 161 color palettes, 57 font pairings, and 99 UX guidelines across 10 stacks.

Query your Lark attendance clock-in records.

Disciplined debugging workflow: reproduce → minimise → hypothesise → instrument → fix → regression-test.

React composition patterns to scale components without boolean prop proliferation.

Create, develop, and deploy full-stack or HTML applications on Feishu's Miaoda platform with local or cloud-based workflows.

Manage OKR cycles, objectives, key results, alignments, and progress tracking in Lark.

Query Azure costs, forecast spending, and optimize to reduce waste.

Create new agent skills with proper structure, progressive disclosure, and bundled resources.

Get a high-level map of code structure and how components fit together.

Ultra-compressed communication mode that cuts token usage ~75% by dropping filler while keeping technical accuracy.

Streamlined brainstorming for design decisions—move from idea to approved direction fast.

Generate and edit images with 11+ AI models (FLUX 2, GPT Image 2, Seedream, Qwen, Wan) via RunComfy CLI.

Swap faces into videos or images via RunComfy CLI, routing to the right model for your intent.

Edit images with Google Nano Banana 2 on RunComfy—batch up to 20 images while preserving subject identity.