
Show the 「which platform pays most」 chart first, then open source: how Wheatfield got 1,100 stars in 4 days
What it is An open-source tool that runs on your own computer. It has Claude Code write copy for seven platforms, and a Chrome extension fills in the upload pages for Bilibili, Douyin, Xiaohongshu, WeChat Channels, YouTube, X, and Instagram (for the four China platforms, you click publish yourself at the last step). Each day it pulls plays and followers, then calculates how many sign-ups and how much revenue each video brought in.
Why it blew up The author is an AI tutorial creator with 6–80k followers across platforms. He first posted a screenshot of a dashboard showing which platform's users pay the most, which pulled in replies asking for the code. 4 days later he launched it with a long post, a 8-minute video and a single install command. The two main posts together got over 1300 million views, and nearly all the stars came from X.
CHow you could follow Don't build another multi-platform publisher. Instead, take its most eye-catching idea, "which platform brought in revenue," and build it as a small tool that connects to payment dashboards. Or make a version for English-speaking TikTok, Reels and Shorts creators. Copy the launch playbook too: show a results screenshot first, then publish the long post, video and proof-it-works screenshots within one day.
- Data (updated daily) · repo created 2026-11-14 · 1,593 stars
GitHub · stars Data 2026-11-17 → 2026-11-20 (4 days, 4 observations). Points before 10-28 come from deeply / archive backfill, not every day has one
Mai Tian (麦田) is an open-source workspace for self-media creators. AI writes copy for seven platforms, a browser extension fills in the upload pages, and it pulls play and follower numbers every day to work out which post and which platform brought in revenue. Plenty of tools already push videos to multiple platforms. It went from 4 to 0 stars in 1,100 days because of its author's own accounts and a tight publishing schedule. Below: what it is, where the stars came from, who can copy this playbook, and why you shouldn't build another "multi-platform publishing tool."
One screenshot, 49 replies
On Nov 2026 12 at 4:20 p.m. Beijing time, X account @czzzzzzJ_ (Maidang mdldm, an AI tutor who calls himself a "former big-tech test developer" with "80k followers across all platforms") posted a screenshot with two lines: "Question: which platform's users are most willing to pay? Answer: it's still Xiaohongshu." The image was his own data dashboard, ranking each video's sign-ups and payments by platform. The post got 37.6k views, 181 likes and 49 replies.
That night at 11:30, he followed up: "Built this for fun and didn't expect everyone to like the workspace. Open-sourcing it soon, would you support it?" It got 87 replies. Two days later he teased: "Mobile widget version is live, open-sourcing it for everyone Monday."
Monday evening, Nov 16, open source. 17:50 He posted a long X article, "The AI self-media workspace everyone was asking about, I just open-sourced it," with 338k views. 6 minutes later he pinned it. The opening line: "woc! Maidang just open-sourced my entire AI self-media distribution and monitoring system!" It came with a 8-minute video and a closing line asking people to repost, star the repo and 🌟. It got 3,004 likes and 650k views. 47 minutes later, a screenshot captioned "20 minutes after launch, friends already got it running" got 668k views (X page data, 11-18). The next day he posted the numbers himself: "Before launch 0 stars. 24 hours later 689 stars and 254 forks, and 1,600 people clicked through to GitHub, mostly from X."
What it actually is
Three pieces work together. It's MIT-licensed and runs on an always-on Mac (the community has already added a Windows install script):
- A skill for Claude Code / Codex: it watches a video, writes titles and copy for seven platforms (Bilibili, Douyin, Xiaohongshu, WeChat Channels, YouTube, X, Instagram), then produces a publishing checklist and validates it (character limits, banned words, cover aspect ratio).
- A Chrome extension, "Maitian Publish Helper" (listed in the store on Nov 14): it opens each platform's upload page in a dedicated Chrome profile, uploads the video and cover, fills in the copy and sets the schedule. For X, YouTube and Instagram it submits directly. For the four Chinese platforms it only fills the form and you click publish yourself.
- The core app (Node + SQLite): manages the publishing plan and prevents duplicates, retrieves links when posts go live, pulls plays and followers from each platform daily, and produces a web dashboard, calendar, per-post review, weekly report and Feishu cards. If you connect your own website, it also gets a "conversion page" that uses UTM tags to show how many sign-ups and how much revenue each video brought in.
The first paragraph of the README is blunt: "One person making content has to publish to six or seven platforms... and still wants to know which post took off, which platform gained followers, and which platform actually brought in payments." The "platform pitfalls" section is full of hands-on detail, for example how WeChat Channels changed a scheduled post after 2 hours, or how Xiaohongshu rejects a finished video with "no video stream info" unless it's processed for faststart. This is a tool its author actually uses, not a demo.
Where the stars came from
| Date | What happened | star |
|---|---|---|
| 11-12 | Screenshot tests demand, promises open source that night | Repo not yet created |
| 11-14 | Repo built, extension listed, Monday open-source teased | 0 |
| 11-16 | Long post + 8-minute video + "20 minutes, someone already got it running" | +127 |
| 11-17 | Posts like "group full" and "24 hours, 689 stars"; community PRs merged | +672 |
| 11-18 a.m. | Growth slows | 1,100(fork 362) |
(Daily numbers from trendshift; the last row is the GitHub page.)
A few details show the people clicking through meant to use it: the fork-to-star ratio is about 33%, while new repos of similar size usually land around 10% (estimate). In 5 days 19 PRs were merged, and by the second day outsiders had contributed Windows support and a one-click install script. Half the issues asked how to join the WeChat group. The README says groups 1, 2 and 3 are full, and now points to group 4. WeChat groups cap at 500 members, so over a thousand people joined in two days (extrapolated from the cap, not verified).
Nearly all the traffic came from the author's X account. A repo less than two months old doesn't make GitHub trending, and I found no coverage on Hacker News, Reddit, Product Hunt or in the media. A few other Chinese AI accounts relayed it, each getting a few hundred views. On launch day the author and several Chinese AI creators announced a community called WayToBot. I couldn't check how many of the 391 reposts came from people sharing with each other.
Another point: his X account had only 3,000 followers before, yet the launch post reached 650k views. The content itself pried open the recommendation algorithm. His 6–80k followers on Bilibili, YouTube, Xiaohongshu, Douyin and WeChat Channels are what made it possible to produce a 8-minute video and post it across every platform the same day.
Why it took off
The hook was a screenshot with money in it. Multi-platform publishing tools usually sell "saves you time." Mai Tian's first appearance sold "which platform's users pay the most." That speaks directly to knowledge-product creators and one-person companies, who care less about views than about where the revenue comes from. Baiting replies with "open-source it" made the launch feel like responding to demand, and built anticipation in.
The launch package was tight. A long post, a video, a pinned post, a proof-it-works screenshot, and numbers the next day, five or six posts within a day or so, each giving the algorithm new material. A one-line curl install, demo.mjs that runs on fake data so you can see it before committing, a 47-second demo GIF, and a three-tier "just look at the data in 30 minutes" setup meant people who saw the post could install it that night. The first "it runs" screenshots then became the next post.
The product does have two things others don't. First, publishing, metrics and conversion sit in the same workspace. Second, for the Chinese platforms it deliberately doesn't click publish for you, doesn't store passwords, and uploads each video to each platform only once. The author devotes a section of the video to "why the extension insists on not clicking publish for you." For creators afraid of account bans, that's more reassuring than a fully automated solution.
It rode a wave. The Chinese AI scene has been enthusiastic lately about "building a one-person-company workflow with Opus plus skills." The first line of the launch post was "I originally used Opus to optimize my own publishing process across 7 platforms." dsh-media-desk, a similar project from the same circle released in early November (it handles topic selection, scripts and covers), also got 1,474 stars, and group members were already using the two together.
To be clear, the features explain only half of it. The same tool without an audience behind it would most likely be just another multi-platform publishing repo with a few dozen stars.
Who it's competing with
| Project | What it is | Scale | How it compares to Mai Tian |
|---|---|---|---|
| social-auto-upload | Python + cookie-based full automation for Douyin, Xiaohongshu, WeChat Channels, TikTok, YouTube and Bilibili | 16.6k stars, since 2023 | Only publishes; also clicks publish automatically on Chinese platforms |
| Postiz | English-language open-source Buffer that uses official APIs | 41.8k stars | Doesn't cover Douyin, Xiaohongshu or WeChat Channels |
| MultiPost | Browser extension that publishes to multiple platforms in one click, mostly image-and-text | 3,900 stars, 6,000 store users | No metrics or conversion tracking |
| MatrixMedia | Electron batch publishing for multi-account matrices | 1,528 star | Built for account matrices, not personal review |
| creatorhub | Multi-platform content monitoring and scraping dashboard | 3,631 star | Focused on watching others, not on publishing your own |
| Commercial publishing clients such as Yixiaoer and Yimei Assistant | Paid publishing clients | Commercial products | Requires hosting your accounts inside the tool, aimed at team matrices |
Looking only at "publish one video to several platforms," Mai Tian has no unique capability, and this category is crowded. From 10 to November, a batch of "one-click multi-platform publish" extensions also appeared in the Chrome store, all with under 100 users. I haven't seen anyone copy Mai Tian directly.
If it were you
C: the open-source launch playbook is worth learning, but building another multi-platform publishing workspace isn't worth it. Open-source options for this are already plentiful. Mai Tian's stars came mostly from its author's audience and timing, and a follower can't get either. What's worth taking away is the playbook and the angle of tying publishing to revenue, not the category itself.
Who it's for: creators who already have their own publishing workflow, and developers with a few thousand to tens of thousands of followers. Or anyone who can find a creator with an audience to co-launch.
Who it's not for: developers with no accounts who want to chase stars by building "yet another one-click multi-poster." It's hard to differentiate on features, and the Chinese platforms' automation risk controls could tighten at any time.
How to do it:
- Don't build an eighth multi-platform publisher. Build the piece nobody has, and the one that works best as a hook: "which platform and which post brought in revenue," as a standalone tool that connects to payment backends like Stripe, Lemon Squeezy, Xiaobao or Knowledge Planet. Leave publishing out, which also keeps you out of the risk-control problem. (Inference: this is Mai Tian's most compelling point, and so far only its conversion page covers it, and it requires wiring up your own site.)
- Or change the audience: build an English version covering "TikTok, Reels, Shorts, Threads and LinkedIn publishing plus conversion review." Postiz handles publishing, but revenue attribution is rare even in the English open-source world (inference, not checked one by one).
- Whichever you pick, copy the launch package: test demand with a results screenshot showing money first, open-source only once people ask, and on launch day release the long post, video, one-line install and "already running" screenshots together. Put a demo GIF and a "try it with fake data first" command on the README's first screen.
- If you want community and repeat users, copy its approach of putting group and tutorial links at the end of the install wizard and in error messages, and make them dismissible.
How to get users: your main channel is your own X account. The Chinese AI scene is currently receptive to "one-person-company workflow" topics. Without an account, find a creator with an audience and have them present your tool as "the workflow I use myself." Mai Tian and dsh-media-desk being used together by group members shows this circle will cross-promote. GitHub trending won't help, since new repos can't get onto it.
When to move: the "Opus + skills for a one-person company" buzz is at its peak in November. Ship before the end of December. Any later and the same audience will be saturated with similar open-source workspaces.
What would mean we misread it: ① two weeks on, Maitian is still adding dozens of stars a day, and similar repos from authors with no audience are also picking up hundreds of stars, which would show the category itself has pull and deserves a higher grade; ② conversely, if Maitian gained only 79 stars on the day 11-18 and drops back to single digits in the days after, the spike was a one-off audience burst and the case for following it is weaker; ③ if Xiaohongshu and Douyin start mass-banning accounts that use automation tools to fill in forms, the window on the China side closes.
Research notes: competitor table, acquisition evidence, full research report
- Competitors
Name Where Launched Scale How it differs from this one dreammis/social-auto-upload Open source 2023-12 16,633 stars, +234 in the last 7 days Full cookie-based automation; also auto-clicks publish on Chinese platforms; publishing only, no metrics or conversion gitroomhq/postiz-app Open source 2023-07 41.8k star English-language open-source Buffer using official APIs; doesn't cover Douyin, Xiaohongshu or WeChat Channels hanliang97/MatrixMedia Open source 2026-03 1,528 stars, +265 in the last 7 days Electron + puppeteer batch publishing for multi-account matrices; built for account matrices, not personal metrics review MultiPost Chrome 2024-11 3.9k stars; 6,000 store users One-click multi-platform publishing extension, mostly image-and-text; no metrics monitoring or conversion tracking Wechatsync / WeChat article sync helper Chrome 2019 6.6k stars; store 10k users Sync articles to Toutiao, Zhihu and similar platforms. No video. 3441293738/creatorhub Open source 2026-07 3,631 stars, +528 in the last 7 days A panel for monitoring, scraping and reposting content across platforms. It's built for watching other people's content. ZJU-REAL/Easel Open source 2026-08 4,746 star An academic-style full-chain social media agent. Heavy on agents, light on data dashboards. Yixiaoer / Yimeizhushou / Rongmeibao Other Years Commercial SaaS; the Yixiaoer browser extension has 335 users Paid, with account hosting and team matrix features. Maitian is free, runs locally, and never stores your passwords. - Where users come from
- Verified Author X account warm-up (screenshot to test demand): 11-12 posted a conversion dashboard screenshot, "Which platform has the strongest willingness to pay → Xiaohongshu", with 181 likes and 37.6k views. The same evening they posted "Open-sourcing soon, would everyone support it?" which got 87 replies.
- Verified Author X launch (video + long post + main post + proof-of-run screenshots): 11-16 long post, 338k views, 1,834 likes; main post 649k views, 3,003 likes, 391 reposts, with a star request at the end; 20 minutes later a proof-of-run post, 667k views; the author says 1,600 people hit GitHub on day one, mostly from X
- Verified 24 hours of posting numbers and community PR: 11-17 posts like "group hit 380 stars", "open source, 689 stars in 24 hours", "X 3k to 10k followers"
- Verified Relayed by Chinese AI-circle alt accounts: @wlzh, Haitun Notes, Dank Daily and others, each post getting a few hundred views.
- Inferred WayToBot creator-group coordinated reposts: on the open-source day, the author and kin佬, 蓝哥, 文哥, 韭二 and others announced the WayToBot community. The repost list wasn't pulled, so the details are unconfirmed.
- Inferred Author's cross-platform video matrix (Bilibili / YouTube / Xiaohongshu / Douyin / WeChat Channels): the README author section and YouTube channel show this episode's video, but view counts weren't retrieved.
- Verified In-product traffic routed to the knowledge site and WeChat group: PR #25 adds UTM-tagged course and group links to the installer wizard, error notifications and conversion page. The README has a direct link to mdldm.club.
Maitian mdldm-maitian: why it got 1,000 stars in 4 days
Data as of 2026-11-18 (Beijing time, midday). Repo: github.com/CzzzzzzJ/mdldm-maitian
Bottom line
This library didn't take off because of its features. Multi-platform distribution already had free options like social-auto-upload (16.6k stars), MatrixMedia and MultiPost. It took off because the author, Maidang mdldm (an AI tutorial creator with 6–80k followers across platforms), first used a conversion dashboard screenshot on X asking "which platform has the strongest willingness to pay?" to draw out demand. 4 days later, with a 8-minute video, an X long post and a one-line install script, he officially open-sourced it. Those two launch posts together drew about 990k views and nearly 4,800 likes, and almost all the stars came from X that same day. Two things set the product apart: first, publishing, data and conversion tracking live in one workspace; second, for domestic platforms it deliberately stops at filling in the form and leaves the final publish click to you. Those two points gave the posts a story to tell, but the growth came from the author's own personal brand and posting rhythm. The open-source release is also a funnel into the mdldm.club annual membership.
What it is
A multi-platform workspace for content creators, running on an always-on Mac (the community has since added Windows), MIT-licensed open source. It has three parts:
- Publishing skill (
skills/maitian-publish): Claude Code / Codex watches the video and writes copy for seven platforms: Bilibili, Douyin, Xiaohongshu, WeChat Channels, YouTube, X and Instagram. It generatespublish.jsonand validates it. - Maitian Publishing Assistant: a Chrome extension, 2026-11-14 on the Chrome Web Store, id
eaaipbilkhibicihiollecmbpneikpnj. The engine code is inCzzzzzzJ/mdldm-maitian-publisher, a repo created 2026-10-29 with 94 stars. The extension opens each platform's upload page in a dedicated Chrome profile, uploads the video and cover, fills in the copy and sets the schedule. X, YouTube and IG are submitted automatically. The four domestic platforms stop once the form is filled, and a person clicks publish. - Maitian core (Node + SQLite): manages distribution schedules and prevents duplicate posts, collects links when posts go live, pulls daily play and follower data, and outputs a web dashboard, calendar, per-episode review, weekly report and Feishu cards. It also includes a "conversion page" that uses UTM tags to link each video to the sign-ups and revenue it brings in. There's also an iPhone widget.
Source: repo README (read via r.jina.ai).
Growth data
| Date | star | fork | Source |
|---|---|---|---|
| 11-14 (repo created) | 0 | 0 | trendshift: 8 PRs merged on the same day, all from the author |
| 11-15 | 0 | 1 | trendshift |
| 11-16 (official open-source day) | +127 | +112 | trendshift |
| 11-17 | +672 | +149 | trendshift; our GitHub repo 11-17 274 → 11-18 966 |
| 11-18 noon | 1,091 | 359 | GitHub API, live lookup |
- The author 11-17 19:18 wrote: "Before launch we had 0 stars; 24 hours later we had 689 stars and 254 forks, and 1,600+ people clicked through to GitHub, most of them coming from X" (X post 2122650510717047029).
- The fork-to-star ratio is about 33%, far above the median for new repos of similar size (about 11%, per the reference in the github repo SCHEMA). That suggests people who came over actually planned to install it, not just bookmark it.
- 5 days, 19 PRs merged (trendshift). External contributors include ajin558, who added Windows support and a one-click install script (PRs #15, #16), and ymstar, who fixed TTY echo in the installer wizard (PR #20). Issues include a request to add Threads (#14) and many people asking what to do when the WeChat group is full (#21, #22, #28).
- Discussion group: the README says "1, 2, 3 groups are full; the group now open is group 4." WeChat groups cap at 500 members, so over two days at least 1,000–1,500 people joined (extrapolated from the group cap, not verified).
- The repo is less than 60 days old, so it doesn't appear on GitHub trending (consistent with what we've observed in the github repo), and it hasn't made the top 25 on trendshift. That means almost none of the traffic comes from within GitHub.
Growth: where users come from
1. The author's own X account: warm-up → launch → data posts, in three stages (confirmed, main channel)
- On 11-12 16:20, they posted a dashboard screenshot with the caption "Question: which platform's users are most willing to pay? Answer: it's still Xiaohongshu." It got 181 likes and 37.6k views (2120793860641280009). The same day, 23:31 followed with "Built this for fun and didn't expect everyone to love the workspace. Open-sourcing it right away, would everyone support it?" which got 132 likes and 87 replies (2120902228328497317). They drew out demand with results first, then promised open source.
- On 11-14, they posted "The phone widget version is live, open-sourcing it for everyone this Monday" (2121561384253366724), and brought it up again with the post "opus + chatcut + Maitian auto-publish" that had spread widely ("全网 6w").
- On 11-16 17:50, they posted an X long post, "The AI content workspace everyone kept asking about, I'm just open-sourcing it": 1,834 likes and 338k views (2122266042491642027). On 17:56 they posted the pinned post "Wow! 麦当 just open-sourced my entire AI content distribution and monitoring system!" with a 8-minute video: 3,003 likes, 391 reposts and 649k views. It ended with "Please repost and star the repo + 🌟" (2122267555133821359). 18:43 then posted "From publish to 20 minutes, a friend already got it running": 1,973 likes and 667k views (2122279446564757793).
- On 11-17 they posted three in a row: "Woke up and the group is already full, 380 stars", "Group members merged Maitian and dsh's workspace together", and "24 hours after open-sourcing, 689 stars." That same evening they posted "Took my X from 3k to 10k followers... the first repo to hit 24k 🌟 within 1 hours" (2122717153124344255).
All the figures above come from live X lookups (X user czzzzzzJ_, 2026-11-18). The account only had 3k followers before, but the launch post got over 600,000 views, which suggests the X algorithm pushed it beyond the author's follower circle and into the recommendation feed of the Chinese AI community.
2. Relays by Chinese AI-circle alt accounts (confirmed, small volume) @wlzh (762 views), @gaoren7716 Haitun Notes (416 views), @dankypower Developer Daily, plus several relays spotted on trendshift along the lines of "No way, they open-sourced it already?" These are all alt accounts and didn't drive volume (x live search "麦田 开源").
3. Blogger mutual-support circle (timing matches, effect unverified) 11-16, the same day it open-sourced, the author and kin佬 (@KinGao476942), 蓝哥 (@0xlangeai), 文哥 (@MaiYangAI), 韭二 (@li9292), Chris (@ChrisHKAI) and others announced the WayToBot community, and 11-17 evening they hosted a Space (the author 11-16 22:54 reposted it). How many of the 391 reposts came from these people, I didn't pull the repost list, so it's unverified.
4. The author's cross-platform presence (partly confirmed) The same episode was also posted on YouTube, Bilibili, Xiaohongshu, Douyin and WeChat Channels (listed in the README author section; the YouTube channel homepage shows "The AI content workspace everyone kept asking about, I'm just open-sourcing it"). Per-platform view counts weren't retrieved (no YouTube API access).
5. Built-in traffic routing (confirmed) The installer wizard's final screen, error notifications and the conversion page shown when no site is connected all include UTM-tagged links to the knowledge site and discussion group (PR #25, "Leave the author's tutorial and group link at the spots where users get stuck (can be turned off)", merged as 11-17). The README author section links directly to the mdldm.club course promotion page.
Channels not found: no presence on HN, Reddit, Product Hunt, English-language media or paid ads. GitHub trending was never reachable anyway because the repo is too new.
Where the money comes from
The repo itself is free. The monetization path is GitHub / X → WeChat discussion groups (4 groups) → the mdldm.club membership, which is "only one annual VIP" (recorded lessons, live sessions, materials, groups). The README says: "How to set up Maitian, how data is collected, and how to connect your own projects are covered in more detail in the course." The membership price isn't listed on the page. The author describes themselves as "former big-company test engineer + AI RPA unicorn → went free", "Coze course reached a million impressions across the web", and "80k followers across all platforms, an AI knowledge site run for a year" (X bio, mdldm.club). Revenue isn't public.
Competitors
| Name | Where | Launched | Scale | How it differs from Maitian |
|---|---|---|---|---|
| dreammis/social-auto-upload | Open source | 2023-12 | 16,633 stars, +234 over the last 7 days (github repo 11-18) | Fully automated Python + cookie uploads to Douyin / Xiaohongshu / WeChat Channels / TikTok / YouTube / Bilibili, with a skill. It only handles publishing, not data or conversion tracking, and it clicks publish automatically on domestic platforms, which carries higher risk-control exposure. |
| gitroomhq/postiz-app | Open source | 2023-07 | 41.8k star | The English-world open-source Buffer. It uses official APIs and targets overseas platforms. It doesn't cover Douyin, Xiaohongshu or WeChat Channels. |
| hanliang97/MatrixMedia | Open source | 2026-03 | 1,528 stars, +265 in the last 7 days | Electron + puppeteer for bulk matrix publishing. It targets multi-account matrices, not per-creator data review. |
| leaperone/MultiPost-Extension | Open source + Chrome | 2024-11 | 3.9k stars; 6,000 store users | Browser extension that publishes to multiple platforms in one click. Mostly image-text content, with no data monitoring. |
| wechatsync / article sync assistant | Open source + Chrome | 2019 / 2020 | 6.6k stars; store 10k users | Syncs articles to Toutiao, Zhihu, Juejin and similar platforms. Not video. |
| 3441293738/creatorhub | Open source | 2026-07 | 3,631 stars, +528 in the last 7 days | A panel for monitoring, scraping and reposting content across platforms. It's built for watching other people's content, not for publishing your own. |
| ZJU-REAL/Easel | Open source | 2026-08 | 4,746 star | An academic-style "full-chain social media agent" that finds trends, writes and publishes. Heavy on agents, light on data dashboards. |
| woniuniuniu/dsh-media-desk | Open source | 2026-11-02 | 1,474 stars, +542 in the last 7 days | The DeepSeek Harness creator workspace from the same circle, for topic selection, scripts and covers. Group members have already been combining it with Maitian, so they're complementary rather than competitors. |
| Yixiaoer / Yimeizhushou / Rongmeibao / New Media Butler | Commercial clients / extensions | Years | Commercial products. The Yixiaoer browser extension has only 335 users (chromestore library). | Paid SaaS with account matrices and team collaboration. You have to hand your accounts over to their servers. Maitian never stores passwords, runs locally and is free. |
| Kuaitui EasyPost, Fenfa Assistant, PublishPort and others | Chrome | 2026-07 to 11 | All under 100 users | A pile of newly appeared one-click multi-post extensions, all of which only handle publishing. |
The space is crowded. Judged purely on "publish one video to multiple platforms," Maitian has no exclusive capability.
Differentiation
- Publishing, data, and conversion in one: competitors almost all stop at publishing. Maitian does two more things: it pulls playback and follower numbers from seven platforms every day and produces daily and weekly reviews; and it uses UTM tags to work out which episode and which platform brought in how many signups and how much revenue. The teaser image is the conversion page itself ("which platform has the strongest willingness to pay → Xiaohongshu"), which hits knowledge-course creators and solo founders right where it hurts.
- Chinese platforms, no auto-publish click: the plugin deliberately doesn't click publish for you. That means it stores no passwords; on Xiaohongshu it only reads rendered content and imitates human actions, and the same video isn't uploaded twice. The author gives "why the plugin insists on not clicking publish for you" its own section in the video. For people afraid of account bans, this is more reassuring than social-auto-upload's fully automatic approach. It's the author's RPA background at work.
- AI-native entry point: the copy is written by a Claude Code / Codex skill, riding the wave of Chinese-language "Opus + skill to build a solo-founder workflow" posts. The lead tweet literally says "used Opus to optimize my distribution workflow across 7 platforms".
- Super low barrier to entry: one-line curl install,
demo.mjsruns on fake data so you can see it working first, a 47-second GIF, a three-level onboarding path ("L1 data-only in 30 minutes"), an iPhone widget, and Feishu cards. Anyone who sees the tweet can have it running that same night, which is why the fork rate is high and people were sending PRs on day one. - A real tool the author uses daily: the README demo uses the real copy and cover from the author's episode 4, and the "platform pitfalls" section is full of hands-on details, like Channels scheduling getting shifted +2 hours and finished videos having to be faststart. That makes it more credible than the typical demo project.
Are these differentiators why it took off? Only half of it. The differentiators (especially the conversion dashboard and "doesn't click publish for you") gave the tweet a strong hook and made people install it and actually want to fork it. But the decisive factor was distribution: the author first tested demand with screenshots, then used a sustained rhythm of "video + long post + lead tweet + proof it runs + 24 hours later, data screenshots," and drove about 990k views on X within Chinese AI circles. The same features under an account with no audience would most likely have been just another multi-platform publishing library with 50 stars.
How many people have seen it
It's circulating in small circles. It's hot on X in Chinese AI circles, and the two launch tweets together got about 990k views; but I found no independent coverage in media, HN, English-language circles, or big accounts, only a few small accounts repeating it. Revenue hasn't been disclosed.
Copycats
Nothing that's a direct copy so far. Several new "publish to multiple platforms at once" Chrome extensions showed up in the Chrome Web Store between 10 and November (EasyPost 10-17, Distribution Assistant 11-17). They all only handle posting, and their users are all under 100, so there's no sign they copied it. A community member combined them with dsh-media-desk, which is a second-hand combination.
What others can borrow
- Show results and test demand first, then open source: a screenshot with "money" in it (willingness to pay, conversion funnel) pulled in comments asking for open source, turning the release into "responding to what people asked for" and building anticipation into the launch itself.
- Launch bundle: a video, a long X post, a lead tweet, a "someone already got it running" post 20 minutes later, and star counts and community PRs shared 24 hours later, feeding the algorithm continuously within a single day.
- Open-source tool as a private-traffic funnel: the README, install wizard, and error messages all carry links to the group and course (can be turned off); when a group fills up, they open the next one, using "the group is full" itself to generate buzz.
- Leave the final step to a human click: on platforms with heavy risk controls, this is a selling point, not a flaw.
- Tie publishing to revenue: what creator tools sell isn't time savings, it's knowing "which platform actually brings in business."
Still unclear
- Out of 391 reposts and nearly 4,800 likes, I couldn't pull the repost list to see how many came from coordinated reposting in creator mutual-aid circles such as WayToBot.
- I couldn't get the view counts for the launch videos on Bilibili, YouTube, Xiaohongshu, and Douyin.
- No data on the price of the knowledge site's annual VIP or the conversions from this wave.
- Real Chrome plugin installs: our chromestore database captured 11-16 users at 0, store user counts are delayed, so actual installs are unknown.
- How long the heat lasts: on day 11-18 it only rose +79 (Trendshift), which suggests a one-off pulse; sustaining it will depend on the author continuing to update and post videos.
Sources
- GitHub repo README, PR and issue lists, commit history (GitHub API, 2026-11-18)
- Our GitHub library repo_day / repo_profile (11-17, 11-18 snapshots, competitor star counts)
- trendshift.io/repositories/361325 (daily stars, forks, PRs)
- Live X pulls: x user czzzzzzJ_, x post 2120793860641280009, x live "麦田 开源" (2026-11-18)
- chromestore database: user counts for the Maitian publishing assistant and similar plugins
- mdldm.club/ai-course-promotion, X profile bio, Jina search results