Apr 01, 2026 11:00:00

RFC 406i is a standard protocol for processing and discarding low-quality AI-generated pull requests.

'AI-generated, low-quality

pull requests ' submitted to source code repositories have become a serious problem for the open-source development community, and some projects have even gone so far as to close pull requests from external contributors . RFC 406i is a standard protocol for handling and discarding such AI-generated, low-quality pull requests.

RFC 406i - The Rejection of Artificially Generated Slop (RAGS)
https://406.fail/

When you access the above URL, you will see the messages 'Reject artificially generated slop' and 'Error 406i: AI slop detected.'

In 2025, the term 'AI slop,' referring to 'low-quality, shoddy products generated by AI,' became extremely popular, and Merriam-Webster, the publisher, selected 'slop' as the best word of the year for 2025.

One of the top international buzzwords of the year has been decided as 'slop' (filth), a word used in phrases like 'AI slop' - GIGAZINE

This is followed by a message criticizing the person who submitted the pull request for the AI slop: 'Your pull request has been redirected to this page because it triggered our automated or manual AI slop defense system. Specifically, a human maintenance worker or senior engineer reviewed your post, let out a deep ontological sigh, immediately closed the socket to your post, and pasted this URL.'

In addition, in the chapter titled 'Diagnostic Analysis,' it states, 'Based on the lexical and structural analysis of your submission, we have concluded that your engineering skills are poor. Therefore, you should reflect on this.' Furthermore, it lists characteristics of AI slop, such as 'unnaturally flattering and mechanical phrasing,' 'extremely confident use of completely fictional APIs,' 'redundant boilerplate text that does not solve any actual problems,' and 'the use of the word ' delve ' in the pull request description without any irony,' and points out that 'your submission also meets these characteristics.'

He also stated, 'Project maintainers, security teams, and community moderators, whether unpaid volunteers or exhausted colleagues, are working under severe resource constraints,' and criticized programmers who create AI slops, saying, 'Project trackers, forums, and repositories are not dumping grounds for unverified copy-and-paste output designed solely to earn green squares on GitHub, or to collect unfounded bug bounties, or to artificially inflate sprint speed, or to maliciously conform to corporate KPI metrics.'

In addition, it stated that to regain the trust lost due to generating low-quality pull requests using AI, the following steps 'must be followed in order' were to be taken.

1. Delete any local branches, text files, or hypothetical vulnerability scripts that generated low-quality AI-generated code using the rm command .
2: Hard reboot your brain.
3. Thoroughly review the actual codebase, project documentation, or threat model to manually verify the status and logic of your work.
4. Do not return to work until you have gained verifiable consciousness and can type with your own human fingers.

As can be seen from the aversion to AI slops that permeates the text, 'RFC 406i' is a joke site disguised as a standard protocol for detecting AI slops, but is actually a satire of AI slops, or rather, of the people who create them.