docs(s2s): sweep stale hydra voice/default claims left after #384 - #385
Open
abhishekmishragithub wants to merge 2 commits into
Open
docs(s2s): sweep stale hydra voice/default claims left after #384#385abhishekmishragithub wants to merge 2 commits into
abhishekmishragithub wants to merge 2 commits into
Conversation
#384 corrected the Hydra model card + AsyncAPI to reflect the wire (hydra-v1.0 as first-class tag, real voice rosters, sterling default, unknown-voice rejection). Four surrounding pages still cited the old wren/sloane/marlowe/reed/knox/tate voices + the false silent-fallback claim + missed hydra-v1.0 as a migration target. Aligning them: - overview.mdx: Note callout now covers all three tags without hardcoded rosters; points at the model card for the per-version voice reference - managing-sessions.mdx: voice field description points at the model card, states sterling default, and corrects the fallback claim to hard error frame with the exact code + param - 2026-08-20-hydra-v1-1-release.mdx changelog: replaces the wren/etc claim with a pointer at the model card + notes hydra-v1.0 is also served and is the server's own named migration target - deprecation-notices.mdx: adds hydra-v1.0 as the migration target the server's warning frame actually names Live-verified end to end: default across all three tags = sterling, unknown voice returns error(invalid_request_error) with param=session.configure.session.voice.
sentinel was missing on the new page; ran scripts/inject_agent_callout.py
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Small cleanup after PR #384. That PR corrected the Hydra model card + AsyncAPI, but four surrounding pages still carried the old claims. Sweeping them so the docs are internally consistent.
What was still stale
Live verification
Reprobed against `wss://api.smallest.ai/waves/v1/s2s` before writing:
Test plan
Discrepancy to flag (not for this PR)
Reprobing surfaced a real gap in #384's roster claim worth a separate discussion. #384 documents `hydra-v1.1` as 10 voices and `hydra-v1.0` as 14 voices with the note that rosters do not overlap. On the wire:
Full accepted set on every tag: `alex, arden, aria, aspen, bellamy, blair, brooks, cole, elena, ellis, emery, grace, harlow, hayes, ivy, juno, kai, lane, lark, leo, linden, marin, maya, noor, pierce, quinn, rowan, sable, sam, sawyer, sterling, vaughn, wilder, zoe`. Ten of these (`aspen, bellamy, harlow, juno, lark, linden, marin, noor, sable, wilder`) do not appear in #384 at all.
This PR does not try to reopen that call. Model card and API ref stay as #384 shipped them. Worth pinging the PM / Aditya so they can decide whether the per-version split is a marketing/quality distinction the server doesn't enforce, or whether the docs should reflect the true 34-voice union.