# "What stumped me was not knowing what this move was called": a comment that became 1,300 stars, motion-prompt · GitHub open-source project

> Source: https://stringers.io/en/github/gh_showheyohtaki__motion-prompt.html (Stringers / 什么值得做, generated 2026-11-08)
> Stringers answers one question every day: what small product can one or two people build in a few months and actually make money from. Numbers come from 15 databases we collect daily, each with source and date.
> For agents: these verdicts and the underlying databases (verified revenue, App Store / Google Play charts, ad libraries, GitHub, Product Hunt…) are queryable live over MCP — tools today / hot / products / directions / direction / data. Setup: https://github.com/komako-workshop/stringers (the user logs in at stringers.io → "配置 agent" to get a personal URL). Site index: https://stringers.io/llms.txt

> The author first posted three "AI command names at a glance" video posts. The top one got 620k views and 13k saves, and it landed the same day as Claude Motion's launch, when everyone was stuck on not being able to name the motions. He turned the posts into an installable library. The first wave of stars came from his own launch post, and the second from a Chinese tutorial and an English big account's long post.

[GitHub open-source project](https://stringers.io/en/github/index.md) · written 11-6

showheyohtaki/motion-prompt · showheyohtaki

# "What stumped me was not knowing what this move was called": a comment that became 1,300 stars, motion-prompt

**What it is:** 192 terms for motion, camera moves, UI, landing pages, and cinematography. Each comes with a moving sample clip, a definition, and an example prompt. Once installed into Claude Code / Codex, you can open the gallery, check the terms you want, and send them to the AI in one click.

**Why it took off:** The author first posted three "AI instruction term list" video posts. The top one got 620k views and 13k bookmarks, and Claude Motion launched the same day, so everyone was stuck trying to name the movements. He turned the posts into an installable library. The first wave of stars came from his own announcement post, the second from Chinese tutorials and long English posts from big accounts.

**B** **How you can follow:** Stop making another motion glossary. Pick a domain where "knowing the name is what lets you give the right instruction" (lighting, sound design, interior styles, typography), follow his format to build 24 nouns + samples + a skill, post the term-list first and see what gets bookmarked, then build and publish the library within a week of a new model launch.

Grade B: the format is proven, other domains and languages are still empty, and it's clear how to copy it. But its stars mostly came from the author's own big audience and the Claude Motion time window. Without an existing audience of your own, it's hard to reproduce that volume. Being passed around in small circles

Data (updated daily) · Repo created 2026-10-14 · 1,372 stars · 7 days +371

GitHub · stars

Data 2026-10-17 → 2026-11-08 (23 days, 23 observations). Points before 10-28 come from deeply / archive backfill, not every day has one

motion-prompt is a glossary of names for instructing AI: camera moves, motion, UI components, landing page effects, and cinematography styles, 192 terms in total. Each term comes with a moving sample clip, a one-line definition, and an example prompt. Once installed into Claude Code or Codex, you can open the gallery, check the terms, and send them to the AI in one click. 23 days, 1,305 stars. The growth came from the author's own 53k-follower Japanese X account, and it coincided with Anthropic launching Claude Motion that week. What followers can take from it is the format: other domains and other languages are still empty.

### One comment

In 2026 Oct 11, Shohei Otaki (@showheyohtaki), CEO of the Japanese IT training school Daytra, posted a video on X titled "For AI instructions: a list of names for video camera moves and transitions." The body had three lines: when making video in Claude or Codex, just saying "make it look nicer" doesn't get you the effect you want; calling moves by name, like "FPV," "dolly zoom," or "whip pan," gets you what you're after; "recommend bookmarking 🔖."

By 11-06, this post had 624k views, 8,371 likes, and 13.3k bookmarks, with 60% more bookmarks than likes. Before and after, he posted two more in the same format: landing page animation names (10-8.2M views, 5,186 bookmarks) and motion graphics names (10-13.1M views, 3,801 bookmarks). All three had more bookmarks than likes, which suggests people saved them to use.

One comment in the thread ended up in the README: what stopped me wasn't imagination, it was not knowing what these movements were called. He said reading that made him want to turn this shared vocabulary for working with AI into a library you can flip through. He created the repo in Oct 14 and announced it the next day.

### What it actually is

Opening the gallery (showheyohtaki.github.io/motion-prompt) shows a wall of small moving cards, each with a number and a name. There are 8 packs of 24 terms each: two packs of motion graphics, two of camera moves and transitions, two of landing page effects, one of UI components, and one of cinematography styles and lighting. Each term has definitions and example prompts in five languages (English, Japanese, Spanish, Simplified Chinese, Traditional Chinese). For example: "title revealed from a mask below, 0.6 seconds, each line slightly later than the last" or "slow push in to her face, move the camera, don't zoom."

What sets it apart from the term lists on X is three ways of using it:

- **Use it as a skill.** After `npx skills add showheyohtaki/motion-prompt` is installed, tell the AI "open the motion-prompt library" and a browser gallery pops up. Click "select," write a line underneath like "use for the opening title," then hit "send to AI." The AI uses only the terms you picked, pulls their specs and reference code, and wires the motion into your project, whether it's Canvas, Remotion, CSS, After Effects, or a video model prompt.

- **Refer by number.** "Make the product video look more professional using #012 and #045" or "write the camera move for #032 as an English prompt for Veo."

- **Make one for your own field.** The make-your-own directory has copy-paste prompts and render scripts that teach a coding AI to translate the set into other languages or swap in terms from your own field. The README points to a few directions: lighting, sound, photography, cooking, architecture.

The usage in the announcement post is four steps: have Claude Code read the repo → say "open the motion-prompt library" → check the terms and hit "send to AI" → say "adapt this for the opening title." That post got 332k views and 5,158 bookmarks, and it linked directly to the GitHub Pages site.

### Why that week

The author's first term post went up in Oct 8, the same day Anthropic launched Claude Motion (generating short animations with code inside Claude). Within days, help docs and coverage came out, and the beta was available on all plans. Lots of people were making animation with AI for the first time and hit the same problem right away: they couldn't say what kind of motion they wanted.

This "AI prompt glossary" trend isn't one guy's doing. He built a page in his repo, found.html, that collects 12 similar posts, most published between 10-06 and 10-13. The English one from @gabriell_lab, 10-06 "If you can't describe scroll animations clearly in Claude or Codex...", got 8,000 likes and 378k views, two days before his. The Japanese scene has a design-term list with 100 picks (772 likes), 6 subtitle-timing styles, 18 camera-move types, and an LP animation dictionary, with posts ranging from a few hundred likes down to a few dozen.

Others stopped at an image or a single video. His account was the biggest (inferred), got the most bookmarks, and he turned the posts into something you can install, check off, and extend. That's why this wave ended up turning into stars for him alone.

### How the stars grew

| Timeline | What happened | star |
|---|---|---|
| 10-08 to 10-13 | Three term-list video posts; 10-08 Claude Motion launched | — |
| 10-14 / 10-15 | Built the library; posted the announcement (332k views) | — |
| 10-17 to 10-22 | Daily +70 to +134 during the first Japanese X wave (peak 10-19) | 162 → 605 |
| 10-23 to 10-29 | Fell back to daily +10 to +20 | 734 |
| 10-30 | Chinese creator @hiwise_aigc recommended "Open-source Skill: Motion Prompt" | 762 |
| 10-31 | A pile of Chinese tutorials: yanchen.app, dpds.me, techlaowang; 11-02 made it onto Sina News | 870(+108) |
| 11-01 | English account @theoxmotion posted a long piece "Make viral videos with Claude Code for free," recommending it as No. 9; 286k views, 3,675 bookmarks | 1,001(+131) |
| 11-02 to 11-06 | +45 to +77 per day | 1,305 |

Data: daily GitHub star history (from 10-17); X post data read as of 11-06. The two jumps on 10-31 and 11-01 and those two posts only line up in time; I can't verify which specific post brought which stars.

The second wave was mostly driven by the Chinese-speaking and English-speaking communities. The community first submitted translation PRs for Simplified and Traditional Chinese, then Spanish and Korean followed. Someone forked a Chinese version (liunnn1994/motion-prompt, 55 stars), 11-05 and someone else created a new -cn mirror. Spanish-language codeia.dev, Korean working-ref.com, and a pile of skill directories also picked it up. It never hit Hacker News or Product Hunt, and mainstream tech media never covered it. The fork-to-star ratio is about 11%, which is high on GitHub, so plenty of people are actually downloading and using it.

### Who it's competing with

| Project | Launched | star | How it differs |
|---|---|---|---|
| pbakaus/impeccable | 2025-11 | 104k | Overall design taste and review commands for coding agents, with no term library featuring sample clips and no camera moves |
| emilkowalski/skills | 2026-03 | 54k | A collection of animation skills from design engineers |
| remotion-dev/skills | 2026-01 | 6,409 | Remotion's official skill, covering the rendering tool layer |
| alchaincyf/huashu-art-motion | 10-06 | 6,316 | Animation skills in 35 art styles; it does styles, not a glossary |
| LottieFiles/motion-design-skill | 2026-03 | 2,834 | Teaches AI animation principles (timing, easing, choreography) rather than a term lookup for humans |
| iart-ai/motion-skills | 2026-06 | 1,248 | 50 skills for "having AI make animation videos" |
| charlie947/motion-graphics-skills | 2026-09 | 576 | Templates for release videos |
| Term-list posts on X | From ¥10-06 | Top single post at 8,000 likes | One-off images and videos; can't be installed or checked off |

Most similar open-source projects teach AI how to do things. This one teaches people how to say what they want. It holds that gap, but only in the area of video and web motion.

### If it were you

**Verdict:** The format is worth copying, but don't copy this domain. Motion and camera moves already have it and a batch of Chinese translations, so another version won't win a spot. To build something, change the domain and audience, but keep the same shape: "terms + samples that move (or can be seen or heard) + an example prompt + a skill you can check and send to the AI."

**Who it suits:** People who are already experts in a field and have a few thousand to tens of thousands of followers: photographers, colorists, audio engineers, interior designers, chefs, illustrators. Ideally, they already post content about "how AI applies to my work."

**Who it's not for:** people with no account of their own who expect the repo to get found through search by itself, or anyone who only half-knows the field and can't name 24 reliable terms. Its star count mostly comes from the author's own 330k views and referrals from others; the content just kept people around.

**How to do it:** pick a field where you need to know the name to write the right prompt, and where AI has only just become capable. Candidates: lighting and set-up (for image/video models), sound effects and music style (for music models), editing rhythm beyond camera language, interior and architectural styles, typography and layout, slide design, food styling and plating shots. Start from the prompts in its make-your-own flow, build 24 entries first, each with a sample and a one-line prompt. A Chinese or Korean version also works; none of these circles has a usable "AI prompt term dictionary" yet (inference).

**How to get people in:** follow its order. First post two or three "AI term list for XX" short videos or long images on your own account, ending with "save this". Watch the saves: if they clearly outnumber likes, then build the repo. The launch post should spell out three or four steps of use and link straight to the gallery. After launch, open translation PRs and ask bloggers in other language circles to pass it on; that's how the second wave came. Add a page to the README listing "similar projects by others" to bring competitors in as allies.

**When to be early:** within a week of a new model or feature launch in your field. Its first post landed the same day as Claude Motion; the next window might be a major update to a music, 3D, or image model.

**What would show you're wrong:** ① your term-list posts get thousands of saves, but two weeks after the repo goes up it has under a hundred or two stars, meaning "saves convert to stars" only worked during this motion-video wave; ② within a month, several same-format repos appear in other fields, none backed by a big account yet each reaching thousands of stars, meaning the format is enough and the account matters less, so move faster; ③ motion-prompt stops updating but still gains 500+ stars in a month, meaning the dictionary itself keeps people, so this part doesn't necessarily need a new field.

Data and sources: GitHub repo page and README (11-06); GitHub daily star history (10-17 to 11-06); X posts from @showheyohtaki (10-08, 10-11, 10-13, 10-15), @gabriell_lab (10-06), @theoxmotion (11-01), data read 11-06; repo found.html; the Claude Motion explanation in the Claude Help Center; Chinese tutorials from yanchen.app, techlaowang.com, dpds.me, Sina News (11-02); codeia.dev, working-ref.com and various skill directories.

[GitHub repo ↗](https://github.com/showheyohtaki/motion-prompt)

**Research notes: competitor table, acquisition evidence, full research report**

Competitors

| Name | Where | Launched | Scale | How it differs from this one |
|---|---|---|---|---|
| pbakaus/impeccable | Open source | 2025-11 | 104k stars | An overall design vocabulary and review commands for agents, not a term dictionary with sample footage; doesn't cover camera moves or video |
| LottieFiles/motion-design-skill | Open source | 2026-03 | 2,834 star | Teaches AI motion principles; doesn't give people terms to look up |
| iart-ai/motion-skills | Open source | 2026-06 | 1,248 star | 50 skills that make AI generate motion videos; no term gallery |
| charlie947/motion-graphics-skills | Open source | 2026-09 | 576 star | A templated skill for making launch videos |
| alchaincyf/huashu-art-motion | Open source | 2026-10-06 | 6,316 star | An art-style animation skill; does styles, not a dictionary |
| UiChemy animation-vocabulary | Web | 2026-10-05 | Not verified | A terminology-mapping skill bundled with a commercial site; no motion gallery and no multi-domain packs |
| Prompt Motion / awesomeclaude motion prompt libraries | Web | Around 2026-10 | Not verified | Shows finished clips and whole prompts to copy; doesn't teach terms |
| Claude Motion | Big-company features | 2026-10-08 | Anthropic's full-plan beta | The generation tool itself; motion-prompt is a complementary instruction dictionary for it |
| Term-list posts on X (@gabriell_lab and others) | Other | From ¥2026-10-06 | Top single post: 8,000 likes / 378k views | One-off images or videos; can't be installed, ticked, or extended into other languages |

Where users come from

- **Verified** **the author's own X account (53k followers) series** : 10-08/10-11/10-13 three term-list video posts got 196k / 624k / 149k views; the 10-15 launch post got 332k views and 5,158 saves, with a GitHub Pages link and four usage steps

- Inference **Platform event: Claude Motion launch** : Anthropic's help docs and press coverage date it 2026-10-08, the same day as the author's first post, bringing many new users who wanted to make motion with Claude

- Inference **Chinese-language bloggers and media reposts** : @hiwise_aigc recommendation 10-30 (reposted by bittide), tutorials from yanchen.app, techlaowang, dpds.me and Sina News 10-28~11-02; corresponds to 10-31 + 108

- **Verified** **English X long-form list** : @theoxmotion 11-01 long post with 286k views, recommending motion-prompt as item 9; corresponds to 11-01 + 131

- **Verified** **Skill directories and community translations** : claudepluginhub, awesomeclaudeplugins, ProSkillsMD candidate listing, agents.nix packaging; Simplified Chinese, Traditional Chinese, Spanish and Korean PRs

## showheyohtaki/motion-prompt: why it works

> Research date 2026-11-06. Data sources: our GitHub repo (from repo_day 2026-10-17), the GitHub API, X post data (read via api.fxtwitter.com), and Exa / Jina / Brave web search.

### One line

A "professional terminology dictionary for prompting AI": motion, camera moves, UI components, landing-page effects, photography styles. 8 packs with 192 terms, each with an animated sample, a one-line definition and an example prompt. It's also a Claude Code / Codex skill: it opens the gallery for people to tick terms, then "send to AI", and the AI pulls specs and reference code to implement them. Author: Shohei Otaki (@showheyohtaki), CEO of Japanese IT training school Dai Trade (デイトラ), with 53k X followers.

### Data

- Built on 2026-10-14, announced on X on 10-15; as of 11-06 it had 1305 stars / 145 forks (data card). The GitHub API showed 1322 that day.

- The curve has two phases: 10-17~10-22 gained +70~134 a day (peak +10-19 on 134), the Japanese X launch wave; 10-23~10-29 dropped to +10~20 a day; from 10-30 it took off again: 10-31 +108, 11-01 +131, then +45~77 a day (github repo_day).

- Last 7 days +543, first 7 days +122.

### Why it took off

#### 1. Demand was validated on X first; the repo is a productized version of a viral post

The author first posted three Japanese "AI prompt term list for XX" video posts on X (all term lists with sample footage):

| Date | Theme | likes | Views | Saves |
|---|---|---|---|---|
| 10-08 | Landing-page animation names | 2,683 | 196k | 5,186 |
| 10-11 | Video camera move and transition names | 8,371 | 624k | 13,280 |
| 10-13 | Motion graphics performance names | 2,185 | 149k | 3,801 |

(Source: api.fxtwitter.com reads of x.com/showheyohtaki/status/2108065874511077598, 2109112248832503895, 2109840044554891464)

Saves were far above likes, showing this is tool-type content people keep to use. The README says a commenter wrote "what stopped me isn't imagination, it's not knowing what this move is called", so he turned it into a library. 10-15 posted the launch (x.com/showheyohtaki/status/2110575893332869584): 2,445 likes, 332k views, 5,158 saves, with the GitHub Pages link and four steps: "have Claude Code read this repo → say 'open the motion-prompt library' → tick → send to AI". The first wave of stars mostly came from here.

#### 2. Timing: the same week Anthropic launched Claude Motion, "making animation with Claude" suddenly became a hot topic

The help docs and multiple press reports for Claude Motion (generating short animations with code in Claude) date it 2026-10-08 (support.claude.com/en/articles/17454997; emergent.sh reported Dashboards and Motion beta available to all plans, 10-09). The author's first post landed on 10-08, just as many new users tried making motion or video with AI for the first time and immediately hit "can't describe the motion I want". That's exactly the problem this repo solves.

#### 3. A "term list for AI prompts" trend was already going on X in Japan, and he had the largest account and the most complete version

The author's own found.html (examples/found.json) collects 12 similar posts, mostly from 10-06~10-13:

- @gabriell_lab (English, 14k followers) 10-06 "If you struggle to describe scroll animations in Claude or Codex…", 8,000 likes / 378k views. That came two days before the author's first post and is one of the starting points of this format in the English-speaking world.

- @Codex_jisen 10-13 design terms 100 picks, 772 likes; @morisuku_kazi subtitle timing 6 methods 442 likes; @yosooi3 camera moves 18 types 361 likes; @revolvtech LP animation dictionary 322 likes; @ane5mama photography techniques 281 likes; dozens more likes.

Everyone else stopped at "one image or one video". He turned it into a browsable gallery + an installable skill + multi-language term packs + a template for making one in your own field, and putting it on GitHub brought the stars.

#### 4. The second takeoff came from referrals in the Chinese-language circles and big English accounts

- 10-30 Chinese blogger @hiwise_aigc recommended "Open-source Skill: Motion Prompt" (the content was reposted on bittide.aicompass.dev; the original post's data wasn't retrieved); 10-29~11-02 a string of Chinese tutorials appeared: yanchen.app (10-29), tempedu.io (10-28), techlaowang.com (11-01), dpds.me (10-31), Sina News (11-02, sina.cn/news/detail/5360794258638372). The community also submitted a Simplified Chinese PR (#1,10-16) and a Traditional Chinese PR (#3,10-19,10-20, merged).

- 11-01 English account @theoxmotion (1,558 followers) posted an X long post, "how to make viral videos with Claude Code for $0 (full guide)", with 1,413 likes, 286k views and 3,675 saves. Item 9 was "motion-prompt's 120 moving samples" (x.com/theoxmotion/status/2116684286837330374).

- Corresponding stars: 10-31 +108, 11-01 +131. The timing lines up, but individual attribution can't be confirmed (unverified).

- Other secondary spread: Spanish codeia.dev (10-17), ABAB News (10-17, "Japanese Developer Releases 120 AI Animation Terminology Library"), Japanese note.com (10-17), Korean working-ref.com (10-26), git2me, thereelstack, agentsuses, claudepluginhub, awesomeclaudeplugins and other skill directories listed it; ProSkillsMD has a candidate listing issue; agents.nix packages it automatically.

#### 5. Why it sticks: not a "picture you glance at and leave", but a tool you can plug straight into your workflow

- Skill: `npx skills add showheyohtaki/motion-prompt`, or install via the Claude Code plugin marketplace; saying "open the motion-prompt library" brings up the gallery, tick → "send to AI", and the agent uses only the selected entries, pulls specs and reference code, and wires them in according to the project's implementation (Canvas / Remotion / CSS / After Effects / video-model prompts).

- Entries are numbered and can be cited directly ("use #012 and #045").

- Has api/v1/catalog.json and a spec txt per entry, for agents to read.

- Fork/star ≈ 11%, with several translation forks (liunnn1994/motion-prompt "Chinese version", 55 stars).

### Competitors

| Name | Where | Launched | Scale | How it differs |
|---|---|---|---|---|
| pbakaus/impeccable | Open source | 2025-11 | 104k stars (+7 in 8,787 days) | "A design vocabulary for coding agents": covers overall design taste and review commands, not individual terms with sample footage; doesn't cover camera moves or video |
| LottieFiles/motion-design-skill | Open source | 2026-03 | 2,834 star | Teaches motion principles (timing, easing, choreography) and how AI should move things; not a term lookup for people |
| iart-ai/motion-skills | Open source | 2026-06 | 1,248 star | 50 skills that generate motion or video, built for "let AI do it"; no term gallery for people to browse |
| charlie947/motion-graphics-skills | Open source | 2026-09 | 576 star | 13 skills for making launch videos, templated output |
| alchaincyf/huashu-art-motion | Open source | 2026-10-06 | 6,316 star | 35 art-style animation skills, Chinese-language circle; does styles, not a dictionary |
| remotion-dev/skills | Open source | 2026-01 | 6,409 star | Remotion's official agent skills; a rendering-tool layer |
| emilkowalski/skills | Open source | 2026-03 | 54k stars | A design-engineer animation skill collection (animate, review-animations) |
| UiChemy animation-vocabulary skill | Web/commercial | 2026-10-05 | Not verified | Maps vague descriptions to motion terms; a skill bundled with a commercial site, with no animation gallery |
| Prompt Motion (prompt-motion.com) / awesomeclaude.ai "17 styles" / ai-x-leaders prompt library | Web | Around 2026-10 | Not verified | Shows finished AI-made clips plus the full prompt. It's copying homework, not learning the terms |
| Claude Motion | Big-company features | 2026-10-08 | Anthropic's full-plan beta | The generation tool itself; motion-prompt is its complementary "instruction dictionary" |
| Glossary posts on X (@gabriell_lab, @Codex_jisen, etc.) | Other (X) | From ¥2026-10-06 | Top single post at 8,000 likes | One-off images/videos: can't be installed, ticked off, or extended to other languages |

### Differentiation

1. **A three-part kit: "term + motion sample + example prompt"** , split and numbered by domain. Similar open-source projects are all about "having AI do it"; this one is about "making people spell out what they want."

2. **Humans and agents share one interface** : people browse and tick clips on the web page, and the agent receives a spec URL and reference code. This "select, then send to AI" loop is something the X glossary posts can't do.

3. **Multilingual, open to contributions, with a make-your-own template** : 5 language glossaries. Community PRs added Simplified Chinese, Traditional Chinese, and Spanish, plus a Korean PR (#4). It also gives a prompt for "have your coding AI build one for your field," turning it from a library into a format.

4. These differences are why it turned X viral posts into GitHub stars. But the real driver of traction was the author's 53k-follower Japanese account, three posts that had already gone viral, and the timing of Claude Motion's launch. The differentiators decided that "the same post got 1,300 stars for him," not whether there was traffic at all.

### Visibility

The small circle has been spreading it: the Japanese X launch post got 330k views, an English big account mentioned it in a long post, several Chinese tutorials and Sina reposts exist, and skill directories commonly list it. But no coverage on HN, Product Hunt, or mainstream tech media (none found). Star count is still in the thousands. It's open source, so there's no revenue.

### Copycats

- Translated forks and mirrors: liunnn1994/motion-prompt (Chinese version, 55 stars), rainman016/motion-prompt-zh, bbylw/motion-prompt-cn (created 11-05).

- Same-format content: 12 X posts indexed by found.html (most predate or coincide with this repo, so they look like the same wave rather than copies of it).

- The author's own spin-off: showheyohtaki/motion-explainer (74 stars, created the same day), prompts for animated explainers of domain terms.

### What others can borrow

- Test demand first with X short-video or image posts, watch the save and like ratio, and only build the repo once it takes off. The launch post gave a "four-step usage" right in it.

- Turn "one-off content" into a skill an agent can call: gallery, ticking, sending to AI, and a machine-readable catalog, so people who bookmark it can actually use it.

- Ride the window of a big company's new feature launch (Claude Motion) with a companion "dictionary/asset" project.

- Open translation contributions so each language community spreads it themselves (the second Chinese surge is an example).

- A "make-your-own" template turns the project into a format. Other domains (lighting, sound, cooking, architecture) haven't been done yet, so you can copy it directly.

### Uncertain

- Which post drove the spikes on 10-31 and 11-01 can only be matched by timing to @hiwise_aigc (10-30) and @theoxmotion (11-01). Causality between the original posts and stars is unverified; @hiwise_aigc's original post data wasn't retrieved.

- Our database has no star counts for 10-14~10-16 (starting from 10-17).

- Actual skill install counts (skills.sh, etc.) were not checked.

- Our X corpus (x-radar) has no related posts at all. All X data comes from single-post reads via fxtwitter.

This article was written by AI from public data and the data we collect daily. Numbers are sourced, and inferences are labeled.