Skill UI

A content creation framework inspired by Chinese AI media creator Kazik (数字生命卡兹克). Covers data-driven storytelling, emotional pacing, self-deprecating humor, AI product reviews, and scalable workflows for producing impactful AI tech articles, videos, and event coverage.

Distill video taste interviews into validated style packs, apply measured cadence to local footage, and export editable EDL/FCPXML timelines with provenance audits. All operations are local, deterministic, and offline.

Simulates Tianxia Bachang, creator of Ghost Blows Out the Light. Write tomb-raiding adventures featuring feng shui, folklore, and legendary characters in a traditional storyteller style with oral, colloquial narration.

A writing skill that applies the creative mindset of poet Xi Murong to craft lyrical, nostalgic, and visually evocative poetry and prose. It centers on life moments, root-based nostalgia, poetry-painting integration, natural imagery, and youth/time reflection. Suitable for love poems, homesickness-themed poems, and artistic prose.

A writing method skill that applies Yu Hua's fiction techniques: suffering narrative, cold tenderness, folk perspective, and time reconstruction. Suitable for crafting poignant characters, first-person storytelling, and realistic prose.

A skill that encapsulates Eileen Chang's fiction writing mindset, including her aesthetics of desolation, contrastive juxtaposition, and urban settings in Shanghai and Hong Kong. Guides writers in crafting character-driven stories with psychological depth and material details.

Compile production-grade visual asset sheets, cinematic storyboards, and reference-based video generation prompts. Handles character identity boards, multi-panel grids, action beats, and R2V/I2V/T2V prompts with consistent identities, camera language, reference responsibilities, and sound design constraints. Intended for creators who need actionable prompts for image and video generation rather than raw editing, model parameter lookup, or evaluating finished videos.

Provides domain knowledge for compiling creative settings into executable image generation prompts for characters, scenes, and props. Covers reference layout modes, composition and camera language, avoidance of image review false positives, and cross-variant identity stability. Used for designing visual assets and writing or fixing VisualVariant and style reference prompt grids.

Simulates real beta readers who meet a draft cold and under time pressure: a skim gate, a no-lookahead timed read with quit points, next-day recall, and a trust ledger of the implied author. Reports where the reading broke so authors decide what to fix before publishing.

Generate and cut videos against a distilled style pack: plan takes, generate with fal, apply measured LUT grading, edit at cadence, weave footage, composite overlays, mint 3D props, and verify results numerically.

Measure reference videos into reusable style packs: 3D LUT grades, cut rhythm, stills, overlay plates, and specs for generative models. Use to capture looks, build repeatable grades, and match pacing.