Adblock4limbo: a per-site ad removal script for Quantumult X, Surge, Loon, Stash and Tampermonkey
Dunai ad-blocking userscript project for Quantumult X, Surge, Shadowrocket, Loon, Stash and Tampermonkey. Uses JS/CSS to strip ads from search engines (Bing/Google), content-farm results, and pop-up/video ads on many video, ACG, novel and manga sites, plus immersive translation and M3U8 extraction, for a clean, ad-free browsing experience.
At a glance
- What is it?
- Adblock4limbo is a JavaScript and CSS rewrite set that strips ads from roughly thirty named sites (search engines, streaming, manga and adult video) through iOS proxy clients or a userscript. It is narrow by design, and the maintenance burden falls on the maintainer, not on you.
- Who is it for?
- Adopt Adblock4limbo if you already run Quantumult X, Surge, Loon or Stash on an iPhone or iPad and you want the specific sites on the maintained list cleaned up, especially video pages where pre-roll ads or M3U8 extraction matter. Do not adopt it if you want general ad blocking across the web, if you browse mainly on a desktop, or if you cannot enable MITM for the target hostnames.
- Can I use it commercially?
- Yes. MIT is a permissive licence: you can use, modify and sell software built on it, as long as you keep its copyright and licence notices.
- Is it still maintained?
- Yes. The repository last received commits 2 days ago.
- What is it written in?
- Mainly JavaScript, according to GitHub's language statistics.
Answers come from the project's GitHub data, last synced on November 15, 2026, and from our analysis. They are not legal advice.
Star history
- limbopro/Adblock4limbo4,582 ★ · +67
| Repository | 2026-09-23 | 2026-11-15 |
|---|---|---|
| limbopro/Adblock4limbo | 4.5k | 4.6k |
67 Stars in the last 53 days, measured from the daily GitHub snapshot. Compare with other projects
Editorial analysis
The problem Adblock4limbo targets: ads that generic filter lists miss
Filter lists work on request URLs and CSS selectors. They are bad at ads that a site's own JavaScript injects after load, at a pre-roll countdown embedded in a player, and at popups that fire when you leave a page. Adblock4limbo goes after that class of ad. The README describes the effect as removing "Javascript 弹窗/Ads横幅" on specific sites, and it lists content-farm results in Google and Bing search pages alongside image ads on streaming and manga sites.
The audience is narrow and the README says so. The project is for iOS browsers (Chrome, Safari and similar) routed through Quantumult X, Surge, Shadowrocket, Loon or Stash, with Tampermonkey users supported "次之或不完美支持" (second, or imperfectly). For desktop browsers the README points readers to uBlock Origin instead, and for Android it suggests Firefox plus uBlock Origin. That is an unusually honest scoping statement, and it tells you the project is not trying to be a general blocker.
One consequence is worth stating plainly: coverage is a list, not a rule engine. The README says about 30 sites are supported and points to Adblock4limbo.weblist as the authoritative, continuously updated source. If your site is not on that list, the project does nothing for you until someone adds it.
How the rewrite and userscript pipeline actually works
The repository ships the same logic in several packagings. Adblock4limbo.js is the rewrite script, Adblock4limbo.user.js is the userscript, and Adblock4limbo.function.js renders the navigation page in the lower right corner. Distribution files sit at the top level: Adblock4limbo.conf, Adblock4limbo.list, Adblock4limbo.module, Adblock4limbo.sgmodule and Adblock4limbo_surge.list, plus per-client directories (QuantumultX/, Surge/, Shadowrocket/, Clash/) and a CSS/ directory holding the selectors.
The mechanism is a response-body rewrite. The proxy client decrypts the HTTPS response for a matching URL, the script runs against the HTML, and it removes ad nodes or injects CSS before the page renders. The README's own example uses script-response-body as the rewrite type, and it stresses that the domain must appear in hostname for MITM to work at all. The README also documents a second function: the script can inject a navigation button and a general selector set described as "1.8万+通用去广告CSS选择器" onto arbitrary pages, not only the maintained sites.
The URL matching is where most misconfiguration happens. The README shows a regex that matches requests containing bbc or ted while excluding .css, .js, .mp4 and .font, written as ^(?=.*(bbc|ted))(?!.*\.(css|js|mp4|font)).*$. Excluding those extensions is not decoration: rewriting a media response as HTML would break playback. The project also splits some work into a separate rule file, since a few sites (Pornhub, 91Porn, 韩剧TV) are marked in the README as needing an accompanying ad-blocking routing rule.
Installing Adblock4limbo in Quantumult X and a first real test
The README's configuration section is written for Quantumult X first, and it says the routing reference should be placed at the top of the routing list. Add this URL as a routing reference:
https://limbopro.com/Adblock4limbo.listAfter adding it, the client fetches the rule set on update. If you place it below other routing rules, those rules may match first and the ad domains will never be routed as intended, which is why the README asks for the top position.
The rewrite side is separate. If you want the script to run on a domain outside the maintained list, you have to declare that domain for MITM and then attach the rewrite script, as the README's example does for BBC News and TED:
hostname = www.bbc.com, www.ted.comWith the hostname declared, add a rewrite of type script-response-body pointing at https://limbopro.com/Adguard/Adblock4limbo.js, and match only the page requests you want with a regex such as ^(?=.*(bbc|ted))(?!.*\.(css|js|mp4|font)).*$. The README warns that the page domain must be present in hostname or the rewrite will not fire.
For a first real check, open one of the maintained sites and look for the navigation button the script injects into the lower right corner of the page. Its presence means the rewrite ran. If ads remain, the README's own troubleshooting order is: clear the browser cache, retry in a private window, and then file an issue with a screenshot and the page URL. It also warns that a browser with a built-in filter, Alook for example, can conflict with the script, and it recommends Chrome over Safari.
Limits: MITM, cache, and sites that drop off the list
The hardest constraint is MITM. Every rewrite depends on the client decrypting the target domain, and the README's FAQ section names www.google.com:80/443 N/A MitM failed as a known failure. If you are unwilling to install a certificate and enable decryption for a domain, this project cannot help you, and no amount of selector work changes that.
Coverage decay is the second limit, and the README is unusually candid about it. Several sites are struck through in the README (低端影视, 泥巴影视, 哔嘀影视, 奈菲影视, 真不卡, 91美剧网, 555影院). The README also asks users to report domain changes, which tells you the list is maintained reactively rather than by crawling. A site that moves today may be broken until someone files an issue.
Third, the userscript path is explicitly second class. The README states that proxy clients are the priority, with Quantumult X adapted first, and that Tampermonkey support is partial. If your only option is a userscript, expect gaps on the sites where the proxy-only features such as routing rules matter.
Finally, adult-content domains make up a large share of the list. That is a deliberate part of the project's scope, not an accident, but it means the tool's usefulness depends heavily on whether those are the sites you visit.
How Adblock4limbo differs from uBlock Origin and browser-level blockers
The README itself points desktop users to uBlock Origin, so the comparison is not adversarial. The difference is where the filtering happens. uBlock Origin runs inside the browser and blocks requests and applies cosmetic filters from lists it maintains or that you subscribe to. It cannot touch traffic before the browser sees it.
Adblock4limbo runs in the proxy layer. That lets it rewrite the HTML body of a response, which is how it can strip a player's pre-roll countdown or extract an M3U8 URL, and it means the same rules apply to any browser on the device, not just one. The cost is the MITM certificate, the client-specific configuration, and a coverage list of roughly thirty sites rather than a general-purpose filter syntax.
A second contrast is maintenance direction. uBlock Origin's lists are community-maintained at scale; Adblock4limbo is maintained by one author who accepts issues and reports adding a site within about 24 hours when the request is worthwhile. That is fast for a single site and slow for a hundred, and it makes the project dependent on the author's continued attention.
Maintenance, licence and what the repository does not promise
The last push to the repository was on 2026-10-31, which is recent, and the repository is not archived. The README states the plan has run since 11.01.2020 and that it will be updated over time to cover more sites. The repository has no published GitHub releases, so updates arrive as file changes in the repository and through the hosted URLs the README points at, which means your client picks up changes when it refreshes the routing reference or the rewrite script.
The upgrade cost is mostly on the maintainer's side, but not entirely. If a site changes domain, you have to notice, and the README's remedy is to file an issue with the domain, the page URL and the ad type. If a site is removed from Adblock4limbo.weblist, the script stops helping there and you will not get a warning.
The project is MIT licensed. For most personal use that is permissive and unremarkable. The thing to check before redistributing or bundling the scripts is the licence file itself, and separately the terms of the sites involved, since the project's purpose is modifying pages you do not own. That is a question for your own counsel, not for this article.
Editorial conclusion
Adopt Adblock4limbo if you already run Quantumult X, Surge, Loon or Stash on an iPhone or iPad and you want the specific sites on the maintained list cleaned up, especially video pages where pre-roll ads or M3U8 extraction matter. Do not adopt it if you want general ad blocking across the web, if you browse mainly on a desktop, or if you cannot enable MITM for the target hostnames. Verify first that your site still appears in Adblock4limbo.weblist, since several sites are struck through in the README, and confirm your client's hostname list covers the domain you care about.
Frequently asked questions
Which apps does Adblock4limbo support?
The README lists Quantumult X, Surge, Shadowrocket, Loon and Stash for iOS, with Tampermonkey userscripts supported second or imperfectly. Quantumult X is described as the first-adapted client. Desktop browsers are directed to uBlock Origin instead.
Why do ads still appear after I configure Adblock4limbo?
The README's order is to clear the browser cache, then retry in a private window, and if ads persist, file an issue with a screenshot and the page link. It also warns that browsers with built-in filters, such as Alook, can conflict with the script.
Does Adblock4limbo work on any website?
No. It targets the roughly 30 sites in Adblock4limbo.weblist, and the README says that list is authoritative and updated continuously. The README does document injecting a navigation page and a general CSS selector set onto arbitrary pages, but that requires you to add the domain to hostname and write your own URL match.
Why does Adblock4limbo need MITM and a hostname entry?
The rewrite runs as script-response-body, so the client has to decrypt the response before the script can modify it. The README states the page domain must be in the hostname list or the rewrite will not apply, and its FAQ names MitM failures such as www.google.com:80/443 N/A MitM failed.
Official sources
Add this badge to your README
If you maintain this project, the badge below links readers to this analysis and shows its maintenance status from the daily GitHub snapshot. Paste the markdown into your README; add ?metric=license or ?metric=stars to the image URL for a different field.
[](https://hysenlabs.com/projects/limbopro-adblock4limbo)Star history chart for your README
[](https://hysenlabs.com/projects/limbopro-adblock4limbo)Was this useful?