[miguel-heygen]

What

overlapping_gsap_tweens now also flags a tl.set() that fires while a tween on the same element and property is still running. The message names the element, the property, the set's time and the tween's window; the fix hint says to let the tween finish before the set fires.

Why

The rule skipped any zero-length window, so a set (zero duration) was never compared with a running tween. That is the shape of a one-word lyric's pop-in (fromTo opacity 0 to 1 over 0.16 s) followed by a hide (set opacity 0) one frame before it ends. Played in order the set wins; seeked, as scrub and export do frame by frame, the tween can win, so the old word stays on screen.

Two tweens with a duration are compared exactly as before. A set at or after the tween's end, at its first frame, on another property, or on another element is not reported.

Testing

fails without the fix: packages/lint/src/rules/gsap.test.ts > "a set that fires inside a running tween on the same property" > "reports a hide that lands in a pop-in's last frames, naming the element and both times"

On the devbox: the lint package's tests pass with the change; with the old gsap.ts the new test fails (1 failed, 235 passed). Linting all 756 HTML files under registry/, skills/ and examples/ gives the same 35 overlapping_gsap_tweens findings before and after, so no new false positive.

Size

A small helper and a rewritten loop in one rule, with its tests; it has no other carrier.

[heygengenesis]

Approving at ea210de5. I found no blocker.

Checked:

• Fails without the fix: with the base gsap.ts (d03bdeae9), the named test fails (1 failed / 235 passed). At head, gsap.test.ts passes 236/236.
• False positives: I linted all 756 HTML files under registry/, skills/ and examples/ before and after the change. Both runs give 35 overlapping_gsap_tweens findings, and none of them is a new set-inside-tween finding, which matches the PR body.
• windowOverlap: two tweens that both have a duration are compared the same way as before. A set is reported only when it lands strictly inside the other tween's window, so a set at the tween's first frame or at its end isn't reported. Two sets are never compared. The overwrite: "auto" and unresolved-target skips still run first. No other rule reports the same case, so nothing is flagged twice.

Nit (not blocking): a set inside a repeat: -1 tween on the same property has running.end === Infinity, so the message reads …to Infinitys and the hint says move the set to Infinitys or later. That case is a real conflict, so the warning itself is correct. You could word the infinite case differently in a follow-up.

[github-actions]

Edit accuracy: accurate 267 (base branch 267), smooth 110 of those

The gate passes.
Smoothness is reported in the artifact, not gated. A case fails only if it fails 2 of 3 runs.

Quarantined, measured but not gated (8)

• seqplainkeys-none-center-r0-nested-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): fail / pass / pass, passes
• seqplainkeys-none-center-r0-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-center-r30-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run
• seqplainkeys-none-pct-r0-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-pct-r30-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run
• seqplainkeys-none-px-r0-nested-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-px-r30-nested-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run
• seqplainkeys-none-px-r30-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run