[miguel-heygen]

What

Release v0.9.22: version bump, releases/v0.9.22.md, and the docs changelog. Merging publishes to npm.

Why

Ships James's preview quality work: Studio scrubbing shows full-quality decoded frames (#7379), and native seeks keep a complete picture instead of an unfinished one (#7378). It also keeps page timers on the film's schedule at every render worker count (#7390). It carries the one other change on main since v0.9.21 (#7384, a test fix).

No visible change

Version bump and changelog only. Each change was captured in its own PR.

Test plan

• bun run release:prepare 0.9.22 --from <v0.9.21 commit> ran on current main (a remote Linux copy without tags): the first run drafted the notes, the summary sentence is by hand, and the second run wrote the version bump. It stopped before tagging; no tag was created or pushed.
• The changed files match the v0.9.21 release commit exactly (the same 21 manifests and changelog), plus releases/v0.9.22.md. Every versioned manifest reads 0.9.22.
• bun run format:check exit 0.
• CI on this PR, including the packed package size budgets.
• Desktop walked with this PR's packed packages (oss-release-gate).

[mintlify]

Preview deployment for your docs. Learn more about Mintlify Previews.

Project	Status	Preview	Updated
hyperframes	🟢 Ready	View Preview	Nov 8, 2026, 1:16 AM

💡 Tip: Enable Automations to automatically generate PRs for you.

[heygengenesis]

Approve at ab800eab.

• Version bump and notes only. The base fed14b016 is the current tip of main. All 20 versioned manifests go from 0.9.21 to 0.9.22, the diff has no other line changes, and no manifest still reads 0.9.21. The changed files match the v0.9.21 release commit, with releases/v0.9.22.md in place of releases/v0.9.21.md.
• The notes match the range. v0.9.21..fed14b016 holds exactly the four commits listed: #7379, #7378, #7390 and #7384. Each SHA and PR link is correct, and docs/changelog.mdx matches releases/v0.9.22.md.

Not a blocker for this PR: @hyperframes/share still returns a 404 on npm, and #7391 was closed in favor of fixing this on the npm side. Unless that npm-side fix lands before the merge, the publish job will very likely end red again on share (ENEEDAUTH) after every other package publishes.