fix(studio): force-reload sdk session after undo/redo bypasses suppress window by vanceingalls · Pull Request #1460 · heygen-com/hyperframes

vanceingalls · 2026-06-15T17:09:23Z

What

Fixes a stale-session bug after undo/redo: after a successful undo, the SDK in-memory document was left on the pre-undo (post-edit) content instead of the reverted content.

Why

writeHistoryFile (called by undo/redo to write the before-content to disk) sets domEditSaveTimestampRef.current = Date.now(), which arms the 2-second self-write suppress window in useSdkSession. That suppress window exists to prevent echo-reloads after SDK cutover writes, but it also swallows the file-change event triggered by the undo write — so the SDK session never reloaded to reflect the reverted file.

How

useSdkSession now returns SdkSessionHandle { session, forceReload } instead of Composition | null
forceReload() bumps reloadToken directly, bypassing the suppress window entirely
useAppHotkeys.applyHistory calls forceReloadSdkSession() after a successful undo/redo that includes the active composition path in result.paths
App.tsx passes activeCompPath and forceReloadSdkSession into useAppHotkeys

Test plan

useSdkSession.test.ts: shouldReloadSdkSession pure function tests (path match, no active comp, no path in payload)
Integration covered by usePersistentEditHistory.test.ts — verifies undo writes correct before-content to disk; the force-reload path is React hook internals not unit-testable without a render environment

…rops onStart, onUpdate, onRepeat were missing from the assertion; only onComplete was checked. All four are listed in DROPPED_VAR_KEYS so all four belong here. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

… (T6e)

- gsapParserAcorn: top-level variable targets now resolved via program-scope null-key fallback in lookupBindingFromAncestors (const el = querySelector...) - gsapParserAcorn: fromTo guard requires args.length >= 3, preventing undefined args[2]/args[3] access when fewer args supplied - gsapWriterAcorn: remove fuzzing fallback in removeAnimationFromScript that silently deleted the wrong animation (from→to ID conversion) - gsapWriterAcorn: valueToCode guards NaN → "0" to avoid broken tween props; safeKey regex aligned to ASCII-only (matching gsapSerialize) - mutate: handleSetGsapTween now includes stagger in extras (was in addGsapTween but missing from setGsapTween) - apply-patches: script case now mirrors stylesheet — op=remove calls setGsapScript("") instead of silently ignoring the patch Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Documents the recast-baseline trust relationship and clarifies that motionPath parity tests live in the Phase 3b commit (PR #1379) since the acorn motionPath parser is also added there. Addresses #1370 R1-N1 (Rames). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…dk-playground

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

…ract suite

- setGsapScript: remove element when newScript="" (fixes undo/redo duplicate-script bug) - parseDeclarations: track quotes so ; inside CSS values (data URIs) doesn't split - handleRemoveGsapKeyframe: guard against duplicate-percentage ambiguity (return EMPTY) - resolveKeyframe: return kfs so callers can check uniqueness - handleSetClassStyle: emit op:"add" (not "replace") when no prior <style> element - FsAdapter listVersions: Number(f.split("_")[0]) — was NaN due to underscore in key - FsAdapter doWrite: split try/catch so appendVersion failure doesn't fire error handlers - FileAdapter playground: add content:"" field to satisfy PersistVersionEntry contract Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

… result.code Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

…rride-set cleanup

… scoped ids Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

Expose the concrete adapter factories so consumers no longer reach into deep adapter paths: - createHeadlessAdapter — no-op PreviewAdapter for agents/CI/SSR (no browser) - createMemoryAdapter — in-memory PersistAdapter for tests/headless open - createFsAdapter (+ FsAdapterOptions) — node fs PersistAdapter for local dev Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

Adds fully-qualified scoped ids for addressing elements inside inlined sub-compositions, so callers can target "hf-HOST/hf-LEAF" unambiguously even when bare hf-ids collide across sub-composition boundaries. Changes: - model.ts: resolveScoped() traverses id segments through nested subtrees; isNewHostBoundary() detects host boundaries (dcf ≠ parent dcf handles outerHTML innerRoot edge case) - types.ts: HyperFramesElement gains scopedId field - document.ts: buildElement carries scopePrefix, propagates childPrefix at host boundaries; buildRoots starts with "" - patches.ts: RFC 6902 escapeIdForPath / decodePathSegment for scoped ids containing "/"; all path builders and pathToKey/keyToPath updated - session.ts: getElement() matches by scopedId; find() returns scopedIds; orphan cleanup decodes RFC 6902 before key comparison, preserves removal markers, purges property sub-keys for both bare and scoped ids - mutate.ts: all element handlers use resolveScoped instead of findById; handleRemoveElement collects full subtree hf-ids before removal for complete GSAP animation cascade (Q3 fix); validateOp uses resolveScoped 20 new contract tests in session.subcomp.test.ts covering resolveScoped, scopedId propagation, dispatch to scoped targets, RFC 6902 patch encoding, override-set key format, orphan purge, and serialize stability. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

Closes the last headless-testable Stage 6 gap (F9 workstream C). `find({ composition: "hf-host" })` returns all scopedIds whose prefix matches the given host id — i.e. every element mounted inside that sub-composition, at any depth. Combinable with other FindQuery fields (tag, text, name, track). 3 new contract tests in session.subcomp.test.ts. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Browser-fetch-based PersistAdapter for Studio REST file API. Uses fetch-only (no node:fs) so it is safe to bundle in Vite. - packages/sdk/src/adapters/http.ts — HttpAdapter class + createHttpAdapter factory - packages/sdk/src/adapters/http.test.ts — 14 unit tests (fetch mocked via vi.stubGlobal) - packages/sdk/src/index.ts — re-export createHttpAdapter + HttpAdapterOptions - packages/sdk/package.json — ./adapters/http subpath export (dev + publishConfig) Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

Add per-path promise queue so rapid successive writes to the same composition file cannot race at the server. Different paths still write concurrently. Two new tests (RED→GREEN verified). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…lush-drains-two test - Add optional headers?: HeadersInit | (() => HeadersInit) to HttpAdapterOptions for cross-origin / CLI / auth injection (function form refreshes on each PUT) - Document listVersions/loadFrom as intentional no-ops (server versioning not exposed) - Add flush-drains-two-concurrent-writes test to mirror fs adapter T13 coverage Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

Adds setSelection(ids: string[]) to Composition interface and CompositionImpl. Fires selectionchange; does not touch undo stack or patch stream. 11 contract tests: get/set/clear, event firing, copy semantics, no undo/patch side-effects. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Skip event dispatch when ids are identical (same length, same order) to prevent double-firing selectionchange from callers that call setSelection with the same list. Two new tests (RED→GREEN verified). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

Reviewer noted it is dead surface in this stack — no caller uses it. Add comment explaining it is wired up in stage 8 when the preview host pushes selection events up to the SDK session. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

Creates useSdkSession hook: fetches active composition HTML, opens an SDK Composition backed by createHttpAdapter, disposes on comp/project change. Session is idle (no dispatch routed yet) — Step 3 wires edit ops through it. Also removes createFsAdapter from SDK main entry (Node-only; subpath-only: @hyperframes/sdk/adapters/fs). Required for Studio typecheck to pass when importing @hyperframes/sdk — fs.ts uses node:fs/promises which Studio's tsconfig does not include. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

useSdkSelectionSync: effect that calls session.setSelection(hfIds) whenever domEditSelection or domEditGroupSelections changes. Maps each entry's hfId; skips entries without one. Pure additive — no existing hook modified. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Build the HttpAdapter first, then call adapter.read(activeCompPath) instead of duplicating URL construction with a raw fetch. Eliminates the /files/encode duplication already in HttpAdapter.read(). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

Wire onDomEditPersisted callback from useDomEditCommits into useDomEditSession, calling reportShadowDispatch (flag-gated via VITE_STUDIO_SDK_SHADOW_ENABLED) to dispatch equivalent SDK ops alongside the server patch path and emit sdk_shadow_dispatch telemetry with mismatch details. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…ismatch Wrap the dispatch loop in try/catch so a throwing SDK dispatch never propagates to Studio UX. Returns dispatched:false with kind="dispatch_error" and the error message for telemetry. One new TDD test (RED→GREEN verified). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…chOperation import Wrap the shadow dispatch loop in session.batch() so a mid-loop throw cannot leave the SDK session in a partially-applied state. Without the batch boundary, one failing op would update some elements but not others, diverging the shadow session from the real one. Rename reportShadowDispatch → runShadowDispatch to eliminate the misleading 'report' prefix — the function mutates the SDK session, it is not read-only. Update the only caller (useDomEditSession). Add missing PatchOperation import to useDomEditCommits (the type was already used in the onDomEditPersisted interface but never imported). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

- http adapter: throw on 5xx instead of silently returning undefined - fs adapter: serialize appendVersion to prevent concurrent pruning race - mutate: GSAP handlers throw (not silent EMPTY) when no GSAP script block - mutate: selectorMatchesId supports scoped hf-HOST/hf-LEAF ids - sdkShadow: fix double data- prefix on attribute patch ops - sdkShadow: call onDomEditPersisted on SDK cutover success path - useSdkSession: fix compRef closure race on dispose; suppress self-write echo - App.tsx: pass domEditSaveTimestampRef to useSdkSession for echo suppression - examples: fix dead !comp.can(op) guard (needs .ok); remove stale id field Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

… review Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…omicity On multi-op payloads, a mid-loop dispatch failure left the SDK session partially mutated. batch() rolls back all ops on throw. Adds: batch-is-called test, multi-op-throw test, GSAP script preservation integration test (linkedom round-trip verified). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…ing TS errors Reviewer noted the 2 s suppress window is a footgun. Add a comment explaining the trade-off (short = echo fires anyway; long = masks real edits) and naming the long-term fix (sequence number / content hash on the persist event). Also fix three pre-existing issues in useDomEditCommits.ts on this branch: - PatchOperation was used in UseDomEditCommitsParams but not imported - onTrySdkPersist was called in persistDomEditOperations but missing from both the interface and the function's destructure pattern - onTrySdkPersist missing from useCallback deps (react-hooks/exhaustive-deps) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

Cherry-pick of af1d91b (dispose race fix) added a second useSdkSession call at the original insertion point from s7step1. s7step3c already had the correct call (with domEditSaveTimestampRef for self-write suppression) at line 156. Remove the duplicate. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

* docs(sdk-playground): accurate README with stage coverage and CanResult * feat(sdk-playground): showcase stage 4 — canUndo/canRedo, removeElement cascade - header undo/redo buttons disabled when canUndo()/canRedo() returns false - History/inspect op section shows live canUndo/canRedo badges (update on every patch) - removeElement logs override-set after removal to demonstrate cascade + orphan cleanup - README: stage 4 row updated to full coverage; Danger section + Ops table annotated - mutate.ts: add fallow-ignore-file code-duplication (structural handler boilerplate) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * feat(sdk-playground): stage 5+6 — adapter exports, scoped ids, find(composition) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * feat(sdk-playground): stage 7 — HTTP adapter + setSelection, update README Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com> --------- Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

…ss window writeHistoryFile arms the 2 s self-write suppress window, so the file-change event for an undo/redo write is swallowed and the SDK in-memory doc stays on pre-undo content. Expose forceReload() from useSdkSession (s7.4) and call it in useAppHotkeys after a successful undo/redo that touched the active composition path. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

vanceingalls · 2026-06-15T17:09:38Z

This stack of pull requests is managed by Graphite. Learn more about stacking.

vanceingalls and others added 30 commits June 14, 2026 21:55

feat(core): acorn GSAP read path with T6b differential corpus tests

0217232

feat(core): acorn GSAP write path — magic-string offset-splice (T6c)

a1ffdfa

feat(core): parse-parity suite for acorn parser (T6d)

18d0380

feat(sdk,core): phase 3b — 8 gsap/label ops + setClassStyle

4da9427

refactor(core): swap studio-api read path from recast to acorn parser…

8ecaeef

… (T6e)

feat(sdk): file-backed fs adapter + setTiming GSAP-script sync; add s…

def5462

…dk-playground

fix(sdk-playground): oxfmt formatting on index.html

281b65a

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

fix(sdk): fs adapter flush() tracks in-flight writes; add to T13 cont…

2929d5a

…ract suite

feat(sdk): can() returns CanResult; T4 dispatch-boundary tests

e5685b3

fix(sdk): export CanResult from package root so callers can switch on…

85f4ba6

… result.code Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

feat(sdk): stage 4 — canUndo/canRedo, removeElement GSAP cascade, ove…

58987f3

…rride-set cleanup

docs(sdk): document cascadeRemoveAnimations bare-id v1 limitation for…

002595b

… scoped ids Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

fix(sdk): de-duplicate ids in setSelection

eb59034

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

fix(studio): flush in-flight http writes before disposing SDK session

d75f15e

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

vanceingalls and others added 13 commits June 15, 2026 00:23

docs(sdk): document persistPath as immutable for session lifetime

d86e97c

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Co-authored-by: Miguel Ángel <miguel07alm@protonmail.com>

feat(studio): stage 7 step 3c — sdk cutover for inline-style ops

8788bf1

feat(studio): stage 7 step 3d — extend sdk cutover to all op types

3f742d9

fix(sdk,studio): 15 code-review correctness fixes from second stage 7…

47db28f

… review Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

vanceingalls changed the title ~~feat(core): acorn GSAP read path with T6b differential corpus tests~~ fix(studio): force-reload sdk session after undo/redo bypasses suppress window Jun 15, 2026

vanceingalls changed the base branch from main to sdk-stage7-studio-s7step3c June 15, 2026 17:10

vanceingalls changed the base branch from sdk-stage7-studio-s7step3c to graphite-base/1460 June 15, 2026 17:58

vanceingalls force-pushed the graphite-base/1460 branch from 859cb40 to 5c8b637 Compare June 15, 2026 17:58

vanceingalls changed the base branch from graphite-base/1460 to main June 15, 2026 17:58

vanceingalls mentioned this pull request Jun 15, 2026

feat(studio): s7.5 — make sdk cutover always-on, delete shadow scaffolding #1462

Open

vanceingalls changed the base branch from main to sdk-stage7-studio-s7step3c June 15, 2026 18:03

vanceingalls changed the base branch from sdk-stage7-studio-s7step3c to graphite-base/1460 June 15, 2026 18:48

vanceingalls force-pushed the graphite-base/1460 branch from 859cb40 to 5c8b637 Compare June 15, 2026 18:48

vanceingalls changed the base branch from graphite-base/1460 to main June 15, 2026 18:48

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix(studio): force-reload sdk session after undo/redo bypasses suppress window#1460

fix(studio): force-reload sdk session after undo/redo bypasses suppress window#1460
vanceingalls wants to merge 45 commits into
mainfrom
sdk-stage7-studio-s7step4

vanceingalls commented Jun 15, 2026 •

edited

Loading

Uh oh!

vanceingalls commented Jun 15, 2026 •

edited

Loading

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

vanceingalls commented Jun 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What

Why

How

Test plan

Uh oh!

vanceingalls commented Jun 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

vanceingalls commented Jun 15, 2026 •

edited

Loading

vanceingalls commented Jun 15, 2026 •

edited

Loading