fix(agent-readiness): aipref attachment draft expired — say so on Content Signals - #131
Merged
Merged
Conversation
…tent Signals draft-ietf-aipref-attach, the document that would bind AI usage preferences to content over HTTP, expired in April 2026 and has not been replaced. draft-ietf-aipref-vocab is still alive. The page implied both halves were converging; split them and name which one a site can act on. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Deploying specification-website with
|
| Latest commit: |
d709c2e
|
| Status: | ✅ Deploy successful! |
| Preview URL: | https://3efbef8d.specification-website.pages.dev |
| Branch Preview URL: | https://fix-aipref-attach-draft-expi.specification-website.pages.dev |
The -04 revision expired 1 May 2026, not April. Add the stronger signal: the attach text has had no edits in the WG's own drafts repository since September 2025, while vocab was being edited for publication in April. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
jdevalk
marked this pull request as ready for review
July 31, 2026 09:43
jdevalk
added a commit
that referenced
this pull request
Jul 31, 2026
…#143) The changelog records when *this spec* changed, not when the world did, so `date` is the day the entry lands on main. Three entries kept the date their PR was authored (28th, 28th, 30th) though all three merged on the 31st. - Re-date tdmrep, mcp-handshake and websub to 2026-07-31, renaming each file so the stem still matches. All three went live under an hour ago, so the GUID churn this causes in the changelog feed is as small as it will ever be. - Add a `changed` entry for ARIA in HTML reaching Recommendation and "Using ARIA" being discontinued (#138). - Add a `changed` entry for the expired aipref attachment draft (#131). That one was merged unlogged; it rewrote what the page says about the topic's standing, which is not a citation repair. - Record the dating rule in CLAUDE.md, including why renaming a live entry is not free: the filename stem is `entry.id`, which is both the `<li>` anchor and the feed's `#`-link. Deliberately not logged: #122 (worked-example callout gained a sentence; the recommendation did not move) and #141 (plumbing, and the hub is not live — fold it into the WebSub entry when D1 is provisioned). Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What changed
One paragraph in the Status of the proposal section of
content-signals, splitting the IETF work into its two halves and saying plainly which one is dead.updatedbumped to2026-07-28.Why now
The page said the IETF and IAB Tech Lab efforts were both ongoing and that "the vocabulary is converging". Half of that is no longer true:
draft-ietf-aipref-vocab— alive. Revision-06, last updated 2026-04-27, WG document on the Proposed Standard track.draft-ietf-aipref-attach— expired and archived. Last revision 2025-10-28, so it lapsed in April 2026. Nothing has replaced it, and the WG's August 2026 milestone for an attachment specification is now at risk.attachis the document that would define how a preference actually binds to content over HTTP — i.e. the thing that would putContent-Signal:inrobots.txton an IETF track. A reader following our page today would reasonably conclude that binding is standards-track. It isn't; right now it rests on the IAB Tech Lab spec and on validator convention.Status
optionalis unchanged and still the honest slot — the page already told readers not to treat this as settled. This is an accuracy fix to how it is unsettled, not a downgrade. No page is being promoted or demoted, and no site capability changes.Changelog — maintainer's call
No changelog entry included. Per
CLAUDE.mdthis sits in the borderline band: it is a real correction to what the spec says about a topic's standing, but it is one paragraph on anoptionalpage, not a substantive rewrite. Log it as achangedentry, or leave it out? Happy to add one before this leaves draft.npm run buildpasses; Prettier and the SKILL.md digest check pass via the pre-commit hook.🤖 Generated with Claude Code