Story

Awesome List Backlinks vs AI Spam: Are They Still Worth It?

A list with 13 stars had 91 open pull requests. Erik Gahner published that count on 5 November 2026. Awesome list backlinks now run into queues like this one.

The post is called “Awesome repositories are no longer awesome.” He gives two reasons. Lists go out of date. And “a lot of these awesome repositories are being flooded with AI slop.”

The stars and the open requests are awesome-free-online-tools. His count, at the time of writing: “13 stars but 91 open pull requests at the time of writing.” The pull requests he looked at suggested “generic vibe-coded tools.”

One user in that post had made at least six of those sites: StatOrigin, ScanChess, Sticker Printer Hub, QuizQuip, LiquidityPools, and DefiLPKit. Gahner wrote that this user “submitted more than 100 pull requests to various awesome lists.” He wrote that some were merged elsewhere. On his own stats repository, he closed the StatOrigin pull request.

He put a limit on the complaint in the same post: “My point is not that these awesome repositories are now all bad. They can still be awesome.”

The post reached the front page of Hacker News the same day. It was posted at 11:16 UTC on 5 November 2026. At a snapshot around 21:05 UTC, the thread stood at about 259 points and about 114 comments. People who maintain lists were in the thread. So were people who still use the lists to find tools.

This story is dated 6 November 2026 in Sydney. Sydney is on Australian Eastern Daylight Time, eleven hours ahead of UTC, so that snapshot is the morning of 6 November here. The point and comment totals will have moved.

Why are awesome lists full of AI-generated pull requests?

anushka_gandhi wrote: “It costs zero effort for a spammer to script 100 PRs.” The same comment calls that “a fast-track to maintainer burnout.”

gokaygurcan has helped maintain awesome-javascript since 2018. He wrote that the amount of low-quality pull requests quadrupled, if not more. He also wrote: “How on earth a 6h old repository with zero user is considered "awesome".”

The same comment tracks the fastest pull request after a repository’s first commit. On his tracking, that gap went from 11 hours, to 10 hours, to 6 hours. Those times are his own record.

hyperupcall, on awesome-bash, wrote: “Almost all of the recent PRs are links to AI-generated projects.” The same comment says: “On many issues, I don’t think I’m responding to a real human.”

LeonB wrote: “for the past few months I’ve received several non-awesome PRs per day.”

zodwick wrote: “I’d like to add a third reason: SEO hacking.”

A July case puts one submitter’s volume on a bot’s tally. Dimantika’s 8 July 2026 post describes awesome-ai-agents-2026. A PR Spam Guard bot flagged a submitter: 100 pull requests in 14 days, 60 of them aimed at awesome lists, and an identical title in 37 repositories. Those three figures are Dimantika’s report of the bot. The submitter said the volume “came from a Python outreach workflow I built earlier this month” and that they would stop “once I have enough dofollow merges.” The workflow and the stopping line are the submitter’s account, as Dimantika reports it.

Gahner called the tools vibe-coded. The July submitter described a workflow that filed the requests. Filing thin tools in bulk belongs with the argument over whether vibe coding is killing indie hacking. kdeldycke, who maintains four awesome lists, described the work on the other side of the queue: “Curation takes time and is like gardening.”

A dev.to guide still sells a listing as “a permanent, dofollow backlink from a high-authority GitHub page.” The July submitter said they would stop once they had enough dofollow merges.

On 6 November 2026, external links in the README on the rendered page for vavkamil/awesome-bugbounty-tools carry rel=“nofollow”. awesome.re is one of those links. A merged entry on that github.com page is a nofollow link. Referral traffic and discovery are what the link can still do. The check is that rendered README. Mirrors and third-party list trackers sit outside it.

Dimantika’s FAQ on the July post says “many GitHub surfaces apply nofollow, and search engines discount low-effort link patterns.”

The guide calls the link dofollow. The README that was checked renders nofollow.

Star gates, bans, and a paused inbox

vavkamil maintains awesome-bugbounty-tools. As of 5 November 2026 the repository had about 6,456 stars, and it was not archived. He wrote: “I had to add a CI check to fail repos with fewer than 50 stars.” He also wrote: “In the past two months, I closed more PRs than I have merged in the past 6 years or so.” That comparison is his own count.

The pull request template on sindresorhus/awesome includes the line “Fully AI-generated pull requests are not accepted.”

On 3 June 2026, awesome-privacy paused issue-based submissions. The note on the discussion: “Due to an influx of spam and AI submissions.”

awesome-npm’s contributing notes set a gate in the repo’s own words: “more than 30 days old and the repo should have at least 40 stars.”

gokaygurcan wrote, in a second comment: “I don’t believe the solution to AI problem is using more AI.”

The same kind of label can land on a person who was formatting a request. In April 2026, sindresorhus/awesome issue 4094 is a contributor disputing a rejection labelled “AI generated” after using Claude Code for formatting.

What the rest of the thread said

Gahner had already said they can still be awesome.

illusive4080 wrote: “Completely disagree. When AI was new I found so many great tools through awesome-* repositories.”

LelouBil wrote: “content curation like this may be more desired than ever.”

xtracto defended awesome-scalability.

dannyfritz07 maintains a niche list. He got a pull request he suspected was written by AI. The corrections in it were right, and he merged it.

Other comments treat the format as an old directory. p-e-w wrote: “They’re a kind of modern version of web directories, and they failed for the same reason the original web directories did.”

meindnoch described the lists as content marketing, and treated that as what they had always been.

davedx wrote that many lists are run by people with a stake in the entries. His example is an author who owns one of the tools on the list.

Stale pages are Gahner’s other reason. He wrote that awesome-R was last updated more than a year ago, and that it links a top-50 packages list “as of 2016.”

How do I get my project added to an awesome list?

Pick a list the project actually belongs on. A nilkick guide on using awesome lists at launch says: “Do not force a fit.”

Dimantika, after the July case, wrote: “A single, honest, non-blast submission that describes what the tool actually is gets reconsidered.”

The gates above are public. vavkamil’s check fails a repository under 50 stars. awesome-npm asks for more than 30 days and at least 40 stars. The sindresorhus/awesome template refuses a pull request that is fully AI-generated. Issue 4094 shows that Claude Code, used for formatting, can be labelled that way. One description of what the tool is, sent once, to a list it fits: that is the request those two notes are talking about.

A listing can send a reader. A buyer is a separate piece of work. How to start indie hacking is a product you own, a price, and a buyer you can reach. Whether you need an audience to indie hack is the neighbouring question. Someone else’s distribution can still sell a product. An awesome list is that kind of room, on the terms the maintainers just spelled out.

Are awesome lists still worth it for indie hackers in 2026?

A list a person still maintains can put a project in front of someone already reading that list. illusive4080 found tools that way. LelouBil thinks the curation may matter more now. dannyfritz07 merged a suspected AI pull request because the corrections were right. Gahner left that case open: they can still be awesome.

The dofollow asset in the dev.to guide is a different claim. On the README checked on 6 November 2026, the external link is nofollow. Referral traffic and discovery are the remainder. A 13-star list with 91 open pull requests is a queue. vavkamil’s star check, the template ban, and the awesome-privacy pause are maintainers answering that queue.

Other distribution channels are moving too. Reddit’s login wall is one channel closing. The Show HN clone wave is Show HN as a distribution channel, crowded with copies. Where the indie hacking crowd actually is is a channel question. The state of indie hacking in 2026 is the wider line: shipping is cheap, and attention is expensive.

Building in public is a log of the product as it gets made. A workflow that files the same request across lists is a different use of the week. The tools indie hackers use are stacks founders have already named in public. An awesome-list line is another submission. On the README checked on 6 November 2026, that kind of external link is nofollow.

A short timeline

  • April 2026. sindresorhus/awesome issue 4094. A contributor disputes a rejection labelled “AI generated” after using Claude Code for formatting.
  • 3 June 2026. awesome-privacy pauses issue-based submissions. The discussion cites an influx of spam and AI submissions.
  • Early July 2026. A PR Spam Guard bot on awesome-ai-agents-2026 flags 100 pull requests in 14 days, 60 of them aimed at awesome lists, with an identical title in 37 repositories. Dimantika writes it up on 8 July 2026. The submitter says they will stop “once I have enough dofollow merges.” That line is the submitter’s.
  • 5 November 2026. Gahner’s post. awesome-free-online-tools at 13 stars and 91 open pull requests, his count at the time of writing. One user, six vibe-coded sites, and more than 100 pull requests, as Gahner tells it. The Hacker News thread is posted at 11:16 UTC and reaches the front page. About 259 points and about 114 comments at a snapshot around 21:05 UTC. He writes that the lists can still be awesome.
  • 6 November 2026. External README links on the rendered page for vavkamil/awesome-bugbounty-tools carry rel=“nofollow”.

Have you submitted a project to an awesome list, and would you still?

FAQ

Are awesome list backlinks dofollow?

On 6 November 2026, external links in the README on the rendered github.com page for vavkamil/awesome-bugbounty-tools carried rel="nofollow". awesome.re is one of those links. A merged entry on that page is a nofollow link. Referral traffic and discovery are what it can still do. A dev.to guide still calls a listing a permanent, dofollow backlink from a high-authority GitHub page. Dimantika's FAQ says many GitHub surfaces apply nofollow, and that search engines discount low-effort link patterns. The check is that rendered README on that date. It covers that page, not mirrors or third-party list trackers.

How do I get my project added to an awesome list?

Match a list the project actually fits. A nilkick launch guide says not to force a fit. Dimantika, writing up a July 2026 spam case, says a single, honest, non-blast submission that describes what the tool actually is gets reconsidered. vavkamil's awesome-bugbounty-tools fails repositories with fewer than 50 stars. awesome-npm asks for a repo more than 30 days old with at least 40 stars. The sindresorhus/awesome pull request template says fully AI-generated pull requests are not accepted. In April 2026, issue 4094 on that repository is a contributor disputing a rejection labelled AI generated after using Claude Code for formatting.

Why are awesome lists full of AI-generated pull requests?

On a 5 November 2026 Hacker News thread, people who watch these queues described a flood. hyperupcall, on awesome-bash, wrote that almost all recent pull requests are links to AI-generated projects. anushka_gandhi wrote that it costs zero effort for a spammer to script 100 pull requests, and called that a fast track to maintainer burnout. gokaygurcan, who has helped maintain awesome-javascript since 2018, wrote that the amount of low-quality pull requests quadrupled, if not more. zodwick called the pattern SEO hacking. On 8 July 2026 Dimantika reported a PR Spam Guard flag on awesome-ai-agents-2026: 100 pull requests in 14 days, 60 aimed at awesome lists, and an identical title in 37 repositories. Those three figures are Dimantika's report of the bot. The submitter said the volume came from a Python outreach workflow built earlier that month, and that they would stop once they had enough dofollow merges. That explanation is the submitter's account, as Dimantika reports it.

Are awesome lists still worth it for indie hackers in 2026?

A list a person still maintains can send a reader who is already browsing it. Erik Gahner wrote on 5 November 2026 that awesome repositories can still be awesome. illusive4080 wrote that they found many great tools through these lists when AI was new. LelouBil wrote that this kind of curation may be more desired than ever. The dofollow promise is a separate claim. External README links on the github.com page checked on 6 November 2026 are nofollow, so the listing works as a referral and a discovery path. Maintainers have answered the flood with star checks, a ban on fully AI-generated pull requests, and, at awesome-privacy, a pause on issue submissions from 3 June 2026. A 13-star list with 91 open pull requests, Gahner's count for awesome-free-online-tools at the time of writing, is a queue. One honest request to a list the project fits is the submission the nilkick and Dimantika notes describe.