One Coder Wrote 75% of MediaCrawler's Commits — for 71 People
One account is credited with roughly 75% of the commits behind MediaCrawler's 61,535 stars, out of 71 people who have contributed at all. The repository has never cut a single tagged release in 1,158 days, and GitHub lists its license simply as "Other" rather than a recognized permissive terms. None of these facts are damning on their own — together they change what a team should check before adopting it.
The ledger
MediaCrawler does what its name says: it scrapes notes, videos, and comments off Xiaohongshu, Douyin, Kuaishou, Bilibili, Weibo, Baidu Tieba, and Zhihu. That's a specific, useful job, and the star count says a lot of people wanted it done. The repository is 96% Python and 4% TypeScript, spans 327 files, was created 1,158 days ago — a little over three years — and last received a push 5 days ago. It is not abandoned. It is, by the commit log, mostly one person's project that 61,535 other accounts starred.
Who's actually writing this
Seventy-one people show up in the contributor list. One of them accounts for roughly 75% of the commits. That's the number this case turns on, and it's worth being precise about what it does and doesn't say. Commit share tells you who typed the code — it does not tell you who reviews pull requests, who understands which subsystem, or what happens to the project's direction if that one account goes quiet. We didn't measure any of those things, and we're not going to guess at them.
What we can say: of the 52 weeks of commit history sampled, eight came back with zero commits recorded. That's consistent with a project driven by one person's schedule rather than a team's rotating cadence — bursts of work followed by quiet stretches, rather than the steadier hum you'd expect from a wider bench of maintainers.
61,535 stars is a crowd. 71 contributors is a small team. A 75% commit share concentrated in one of them is the gap between how popular this project looks and how many hands are actually holding it up — and that gap is the thing to weigh before you build production infrastructure on top of it.
No paper trail
Three years, 61,535 stars, and 188 open issues — and not one tagged release. Zero. That means there is no version number to pin in a requirements file, no changelog entry to point to when something breaks, no "known good" tag to roll back to. Anyone consuming this as a dependency is pointing at a moving branch, not a fixed release, whether they realize it or not.
The repository does carry a tests directory and a CI configuration — both present in the file tree as of this check. That tells you a pipeline exists and a test suite exists. It doesn't tell you the suite passes right now, what it covers, or whether the pipeline is green today; we didn't check build status, and this piece isn't going to pretend otherwise. There's also no Dockerfile in the tree, so if the plan is to run this in a container, that packaging work is on you.
The fine print
GitHub classifies the license as "Other" — not MIT, not Apache-2.0, not BSD, not any of the templates a legal team can recognize on sight. That single word is the whole finding here: there's no shortcut to "it's permissive, don't worry about it." Whatever this project actually permits and requires is written into its own license text, and that text is the only source that matters. A team weighing this for anything beyond a personal script needs someone to actually open that file and read it before code from this repository ships anywhere near a product.
Questions people ask
Is MediaCrawler actively maintained?
Yes, by activity: the last push was 5 days before this check on August 11, 2026, and the sampled commit history shows regular, if uneven, activity across the past year.
Who actually writes MediaCrawler's code?
71 accounts have contributed, but one of them is credited with roughly 75% of the commits, as measured via the GitHub API on August 11, 2026.
What license does MediaCrawler use?
GitHub lists it as "Other" rather than a standard recognized license such as MIT or Apache-2.0, which means its actual terms have to be read directly rather than assumed.
Has MediaCrawler ever published a versioned release?
No. As of August 11, 2026, the repository shows zero tagged releases despite being created 1,158 days ago and accumulating 61,535 stars.
None of this makes MediaCrawler a bad piece of software — 61,535 stars and a push five days old say the opposite. But a star count measures attention, not organizational depth, and this repository's own numbers show a wide gap between the two. A hobbyist pulling data for a side project has little to lose here. A team putting this into a production pipeline is taking on an unpinned dependency, maintained mostly by one contributor, under a license nobody outside this repository has written a name for yet.
Before you commit to it: read the actual license file, not the GitHub badge. Vendor or pin a specific commit hash, since there's no release tag to lean on. And go in knowing that if the top contributor's pace changes, so does the project's.
Read the full article
Comments
Post a Comment