[kixelated]

Deletes opus-stalled-share: its own exit condition ("if the rows pass, delete this quest") is met.

Evidence

Scheduled nightlies after moq-dev/moq#8058 merged (2026-11-09T02:20Z), Chromium step reports "within budget" both nights:

row	42659882487 (11-09)	42842827718 (11-10)
opus-48000-bursty	0.009	0.005
opus-48000-wide	0	0.005
opus-48000-step	0.009	0.009
mild / near-zero / fixed-250	0	0

Ceilings are 0.01, except opus-48000-step at 0.022. Both nights ran before moq-dev/moq#8459 merged (2026-11-10T23:29Z); that PR regrades lag and trims, not stalled_share, so it does not change this verdict. The Audio quality job still fails on other steps (native target_ms_p95, push-to-talk), outside this quest.

Choices made without asking

• Graded on two nights rather than waiting for more: the quest asks only whether the rows pass. Bursty sits at 0.009 against 0.01, close to its ceiling; Calibrate re-measures every ceiling anyway.

Review

Skipped at head 4076330a669b2234cd97107c105e7b1aeb93fb11: the diff only deletes a quest file and its links, plus a merge of origin/main resolved by taking main's epic README (it already dropped ## Required). quest check and just check pass. Pullfrog's ceiling note is fixed in this description; the OpenAI review had no findings.

Public API: none. Wire: none.

🤖 Generated with Claude Code

Implementer: Claude Code, claude-opus-5-5, low effort (iterate, XS)

(Written by Claude Opus 5.5 with low)

[coderabbitai]

No actionable comments were generated in the recent review.

Walkthrough

The Chromium Opus stalled-share quest document was deleted. Its entry in the Required list and its link in the Related list were removed from the audio jitter target documentation.

Merge Risk: Minimal · up to 48a5e

This is a documentation-only removal of a quest whose exit condition is reported as met. It has no runtime impact and is ready to merge.

Pre-merge checks: 4

Check name	Status	Explanation
Title check		The title clearly identifies the quest deletion and the reported reason: Chromium Opus rows passed after #8058.
Linked Issues check		Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check		Check skipped because no linked issues were found for this pull request.
Description check		The description directly explains the deletion of the opus-stalled-share quest and provides the reported evidence that its exit condition was met.

Finishing Touches

Simplify code

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

Share

• X
• Mastodon
• Reddit
• LinkedIn

Recent review info

Run configuration

• Source mode: Auto
• Repository: Unavailable
• Branch: Unavailable
• File: No configuration file from this repository applied
• Commit: Unavailable
• Layers: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 98f91647-bb6f-4766-83fe-50fe41dc67c0

Commits

Reviewing files that changed from the base of the PR and between 88572c5 and 48a5e4a.

Files selected for processing: 3

• quest/m1/audio-jitter-target/README.md
• quest/m1/audio-jitter-target/calibrate.md
• quest/m1/audio-jitter-target/opus-stalled-share.md

Files with no reviewable changes: 3

• quest/m1/audio-jitter-target/calibrate.md
• quest/m1/audio-jitter-target/README.md
• quest/m1/audio-jitter-target/opus-stalled-share.md

Included review availability: This review used your included allowance. Your plan provides up to 4 included reviews per hour; 0 remain after this review.

[kixelated]

Automated review by review (OpenAI)

Reviewed commit 48a5e4a.

Overall direction: No actionable findings. Removing the quest matches its explicit exit condition: the two cited scheduled-run logs both show Chromium Opus rows within budget, with the stated stalled-share values. The separate calibration quest remains, and this changes neither player behavior nor budgets.

Verification limits: GitHub-only diff and existing Actions-log inspection. No new workflows, builds, or tests run; two nights do not establish a long-term flake rate.

[pullfrog]

ℹ️ Minor suggestion only. The deletion is sound and I reproduced the evidence; one figure in the PR description is off.

Reviewed changes

Reviewed the complete diff (3 files, 0 additions, 36 deletions): a completed quest is removed along with both of its inbound links.

• Quest deleted — quest/m1/audio-jitter-target/opus-stalled-share.md, after its own exit condition (Chromium Opus rows pass a stalled_share budget) was met.
• Inbound links removed — the epic's Required entry in quest/m1/audio-jitter-target/README.md and the Related entry in quest/m1/audio-jitter-target/calibrate.md.

I verified the evidence independently. Both cited nightlies exist and postdate #8058's merge: run 42659882487 (2026-11-09T16:57Z, head b3fbce18) and 42842827718 (2026-11-10T15:17Z, head cd560bde). Both ran just test audio-quality --runtimes chromium --enforce, printed within budget for the Chromium step, and their stalled column matches the table. #8459 leaves stalledShare in analyze.ts untouched (window.filter((s) => s.stalled).length / window.length), so the pre-#8459 runs grade the same metric. grep finds no tracked reference to the deleted path left, and the Quest check run on 48a5e4a succeeded.

ℹ️ opus-48000-step is budgeted at 0.022, not 0.01

The evidence table says "Ceiling is 0.01" for every row, and the Choices section says "Step sits at 0.009, close to the ceiling". budgets.json gives chromium/opus/step stalled_share: 0.022, at both graded SHAs as well as on main, so that row passes with more than 2x headroom. The verdict is unchanged (every row is under its own ceiling), but the description overstates how tight step is — bursty at 0.009 against 0.01 is the row that is actually close.

Technical details

# Correct the stalled_share ceiling reported for the step row

## Affected sites
- PR description, Evidence table footer — "Ceiling is 0.01" applies to five of the six rows; `chromium`/`opus`/`step` is 0.022.
- PR description, "Choices made without asking" — "Step sits at 0.009, close to the ceiling" is false at 0.022; the near-ceiling row is `bursty` (0.009 vs 0.01).

## Required outcome
- The description states each graded row's own ceiling, or notes that `step` is budgeted at 0.022, so a later audit does not read `step` as near-failing.

## Suggested approach (optional)
- Change the table footer to "Ceilings are 0.01, except `opus-48000-step` at 0.022" and reference `bursty` in the Choices bullet. `test/audio-quality/budgets.json` is the authority; the ceilings were identical at `b3fbce18` and `cd560bde`.

  ^{｜ Fix it ➔ ｜ View workflow run ｜ Using deepseek-v4.1-flash-1017 (free via Pullfrog for OSS) ｜ 𝕏}

[pullfrog]

✅ No new issues found.

Reviewed changes

Reviewed the delta since the last Pullfrog review (48a5e4a): a merge of origin/main and the resulting narrowing of the diff.

• Merged origin/main (529fa6ac9) — main had already dropped the epic README's ## Required and ## Related sections, so the merge kept main's README untouched and the PR is now a two-file deletion (0 additions, 35 deletions).
• Deletion still consistent — calibrate.md on main still carried the Chromium Opus stalled share link this PR removes, and grep -rn opus-stalled-share quest/ finds no dangling reference at 4076330.
• Prior finding addressed — the description now reads "Ceilings are 0.01, except opus-48000-step at 0.022" and names bursty as the row near its ceiling, matching test/audio-quality/budgets.json.

  ^{｜ View workflow run ｜ Using deepseek-v4.1-flash-1017 (free via Pullfrog for OSS) ｜ 𝕏}