MiniMax-AI/mcode-stack / stacks/jiachen/skills/making-content
making-content
VerifiedContent production in three modes: rewrite, Xiaohongshu (小红书) image post, short video. Rewrite (default): turns an AI-sounding Chinese or mixed Chinese-English draft (post, newsletter, engineering blog, commentary) into the author's restrained style: few emoji, short paragraphs, hedged claims, neutral wording, English terms split by spaces. Triggers: 按我的风格改, 去 AI 味, 润色, 改改这篇, or a vague 帮我改改 when the context is a structured draft. Also for a Xiaohongshu image post from scratch (title, cover, cards, body, hooks); read references/xhs-post.md first. Also turns a document, research, notes or an article into a 60-180 s vertical video (抖音, 小红书, 视频号, B站: real sources, voice-over, captions, cover). Triggers: 做成视频, 做个短视频, 剪个 90 秒版本, 配音, 换成我的声音, or complaints of 字太密, AI 味太重, 容易走神; read references/video-pipeline.md first.
| File | Kind | Size |
|---|---|---|
| .gitignore | other | 167 B |
| SKILL.md | skill md | 25.6 KB |
| evals/evals.json | config | 4.5 KB |
| references/layouts.md | markdown | 8.5 KB |
| references/text-density-lint.md | markdown | 4.2 KB |
| references/video-pipeline.md | markdown | 12.5 KB |
| references/voice-and-music.md | markdown | 5.5 KB |
| references/xhs-post.md | markdown | 9.0 KB |
| requirements.txt | manifest | 233 B |
| scripts/capture.py | script | 6.4 KB |
| scripts/compositor.html | other | 27.8 KB |
| scripts/fetch_assets.py | script | 4.0 KB |
| scripts/lint.py | script | 8.6 KB |
| scripts/render.py | script | 4.6 KB |
| scripts/timeline.py | script | 8.7 KB |
| scripts/tts.py | script | 5.5 KB |
| templates/cover.html | other | 2.9 KB |
| templates/script.example.json | config | 5.0 KB |
| templates/sources.example.json | config | 58 B |
references/text-density-lint.md
View on GitHub# Text-density lint
## Contents
- What lint.py measures
- Rules and where the numbers come from
- Fixing violations
- Changing a rule
## What lint.py measures
`lint.py` loads the project's `compositor.html` with the real timeline and captures, renders every shot twice (about 1 s in, and near its settled end) and measures each visible text node: box, rendered font size including transforms, line count, reading load. Then it renders the middle of every caption cue and checks the caption on its own.
Reading load counts 汉字 as 1 and Latin characters as 0.5; punctuation is free.
Two things are not counted as text to read:
- text at 35% opacity or less is texture (code context lines, owner and file path, watermarks);
- everything inside an element with a `data-lint-group` attribute counts as one block (a diagram, a ladder).
## Rules and where the numbers come from
Frame: 1080×1920. Every threshold is in `RULES` at the top of `scripts/lint.py`.
| Rule | Value | Source or reason |
|---|---|---|
| Safe zone | x 60–1020 above y 850, x 60–960 below it; y 180–1600 | 抖音/TikTok overlays: top tabs, the like/comment/share rail on the lower right, the name/description/music block at the bottom [2][3][4] |
| Caption position | top at y ≥ 1380 | low and out of the picture's centre, still above the bottom overlay [3][4] |
| Caption length | 1 line, at most 16 字 | Netflix Simplified Chinese: 16 characters per line [1]; short video uses one line |
| Caption speed | at most 9 字/s, at least 0.6 s on screen | Netflix adult reading speed of 9 characters per second [1]; shorter captions flicker |
| Caption clearance | 60 px free above the caption | the caption must not read as part of the picture |
| Body text size | at least 34 px | legible on a phone at arm's length; labels may be 26 px if at most 24 字 |
| Text blocks per screen | at most 4, caption excluded | one focal point, few eye jumps |
| Reading load per screen | at most 36, caption excluded | a 3–6 s shot is read once, while the voice is talking |
| Font-size levels | at most 3 (sizes within 15% count as one) | clear hierarchy |
| Line length | at most 14 per line | large type, short lines |
| Orphans | last line at least 2.6 em wide | no one- or two-character last lines |
Sources:
1. Netflix, Simplified Chinese Timed Text Style Guide — https://partnerhelp.netflixstudios.com/hc/en-us/articles/215986007-Chinese-Simplified-Timed-Text-Style-Guide
2. PostPlanify, TikTok safe zone checker — https://postplanify.com/tools/tiktok-safe-zone-checker
3. BlitzCut, safe zones for YouTube Shorts, Reels and TikTok — https://blitzcutai.com/blog/safe-zones-youtube-shorts-reels-tiktok
4. Reap, short-form video safe zones — https://reap.video/blog/short-form-video-safe-zones
## Fixing violations
| lint says | Usual fix |
|---|---|
| `N text blocks > 4` | Join label and value into one element; drop a label the voice already says; wrap a diagram in `data-lint-group` |
| `reading load N > 36` | Cut words, since the voice carries the detail; or split into two shots |
| `N font-size levels > 3` | Reuse an existing size, or dim the least important text to texture |
| `too small Npx` | Raise it to 34 px, or shorten it to a label of at most 24 字 |
| `outside safe zone` | Move it inside; bleed layouts may bleed backgrounds, never text |
| `orphan last line` | Break the line yourself with `\n` in `zh`, `zh1`, `hook` or `title` |
| `line too long` | Add `\n` or shorten |
| `caption N chars` | Add a `/` in `cap` |
| `caption N chars/s` or `caption shown 0.xx s` | Fewer 字 in that piece, or move the `/` so the merge pass can join it with a neighbour |
| `caption wraps to 2 lines` | A shorter piece; Latin words are wide |
| `caption too high` / `too low` | Something overrides `#cap`; the caption position itself stays fixed |
| `text crowding the caption` | Move the element up, or dim it to texture; keep the 60 px gap |
After fixing, re-run `render.py --stills` and look at the sheet again: lint catches density, not ugliness.
## Changing a rule
Change a value only for a reason that applies to the whole video, such as another platform's overlay layout or another frame size, keep its comment in `RULES` accurate, and re-lint every shot.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deployment | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |
| xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved. | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |
| shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |