Multi-image NewAPIWG generations were sending base64 reference images
inside the /api/generate JSON body. Two ordinary canvas references can
push that payload near the dev/server body boundary, leaving the route
to parse an incomplete JSON string and surface a raw syntax error.
Move large NewAPIWG references through the existing in-memory image
serving path: the browser uploads large data URLs as multipart image
files, sends short /api/images/{id} URLs in the generation request, and
the route resolves those temporary URLs back to server-local image data
before calling providers. The route also turns malformed/truncated JSON
into a readable validation error.
Constraint: Keep provider payload semantics unchanged after /api/generate receives the request
Constraint: Do not rely on request origin matching because Next dev can report 0.0.0.0 while the browser uses localhost
Rejected: Increase body size limits only | still sends multi-MB JSON and does not protect hosted/proxied environments
Rejected: Compress reference images client-side | changes model input fidelity and still leaves large JSON possible
Confidence: high
Scope-risk: moderate
Directive: Do not forward /api/images/{id} temporary URLs to external providers; resolve them server-side first
Tested: npm run test:run -- src/store/execution/__tests__/nanoBananaExecutor.test.ts src/app/api/generate/__tests__/route.test.ts
Tested: npm run build
Not-tested: Live NewAPIWG generation with the user's exact two images
The welcome screen now matches the Popi.TV brand and gives unauthenticated users two explicit paths: jump to the main-site login page or provide a Popi/NewAPI gateway key directly. The manually provided key is persisted in the local user store and forwarded through the NewAPI-backed generation, model, LLM, and chat routes without requiring a login token.
Constraint: Users need a usable local path when they are not logged into the main site.\nRejected: Keep the startup login modal for missing tokens | it blocked the in-welcome login/key choices.\nRejected: Rename internal workflow prompts and storage keys | those are compatibility/internal identifiers, not the visible brand surface.\nConfidence: high\nScope-risk: moderate\nDirective: Keep manual gateway-key access scoped to NewAPI-backed paths; non-NewAPI providers should continue requiring a login token.\nTested: npm run test:run -- src/app/api/models/__tests__/route.test.ts src/app/api/models/[modelId]/__tests__/route.test.ts src/app/api/llm/__tests__/route.test.ts src/app/api/chat/__tests__/route.test.ts src/app/api/generate/__tests__/route.test.ts\nTested: npm run test:run -- src/components/__tests__/QuickstartInitialView.test.tsx src/components/__tests__/WelcomeModal.test.tsx\nTested: npm run build\nTested: Browser verification for Popi.TV welcome title and two unauthenticated entry paths\nTested: git diff --cached --check\nNot-tested: Full test suite\nNot-tested: npm run lint still fails because the existing Next 16 next lint script resolves /lint as a project directory
P0/P1 reports clustered around unavailable LLM defaults, video media connections, and missing controls on the generation surface. This routes the default NewApiWG LLM away from the unavailable Doubao channel, makes video a first-class schema and canvas connection type, carries request ids into node errors, and fills the composer, trim, audio, and 3D gaps without changing the canvas workflow model.
Constraint: NewApiWG Doubao channels are not reliably routed in the default gateway
Constraint: Video model schemas expose video fields that must not flow through image handles
Rejected: Continue accepting video through image handles | hides incorrect graph types and loses schema-specific dynamic inputs
Confidence: high
Scope-risk: moderate
Directive: Keep media handle types explicit; add schema mapping when adding a new connectable media type
Tested: npm run test:run -- src/store/execution/__tests__/generateVideoExecutor.test.ts src/store/utils/__tests__/connectedInputs.test.ts src/store/execution/__tests__/generateAudioExecutor.test.ts src/store/execution/__tests__/llmGenerateExecutor.test.ts src/store/__tests__/workflowStore.integration.test.ts src/lib/__tests__/llmModels.test.ts src/store/utils/__tests__/nodeDefaults.test.ts src/components/__tests__/GenerationComposer.test.tsx src/app/api/chat/__tests__/route.test.ts src/app/api/llm/__tests__/route.test.ts src/app/api/models/[modelId]/__tests__/route.test.ts src/app/api/generate/providers/__tests__/newapiwg.test.ts
Tested: npm run build
Tested: git diff --check
Not-tested: npm run lint | package script calls removed Next 16 next lint command
Not-tested: npx tsc --noEmit | existing test type errors remain; changed runtime files filtered clean
Image generation settings arrive at /api/generate as top-level aspectRatio and resolution fields, but the NewApiWG provider path only forwarded model-specific parameters. Native NewApiWG Gemini image generation therefore fell back to default imageConfig values and ignored composer/control-panel ratio choices. The route now merges top-level image settings into provider parameters while preserving explicit model parameter overrides.
Constraint: NewApiWG provider implementations read aspect ratio and image size from GenerationInput.parameters.
Rejected: Parse aspect ratio from prompt text | UI settings should be authoritative and prompts are ambiguous.
Confidence: high
Scope-risk: narrow
Directive: Keep top-level GenerateRequest image settings and provider parameters in sync for image providers.
Tested: npm test -- --run src/app/api/generate/__tests__/route.test.ts src/store/execution/__tests__/nanoBananaExecutor.test.ts src/components/__tests__/GenerationComposer.test.tsx
Tested: npx tsc --noEmit --pretty false filtered to generate route files
Not-tested: Full tsc due existing test mock overload errors in route.test.ts
The bottom composer now edits and reruns the current generation node while canvas-visible actions own downstream node creation. NewApiWG media normalization, video frame extraction, and node quick-add keep generated media flows explicit and connected rather than hidden behind the prompt box.
Constraint: Composer must not call providers directly or create hidden downstream branches when a node is selected
Constraint: NewApiWG video families use inconsistent media schemas, so common media aliases are normalized in the provider adapter
Rejected: Let the bottom composer create next-step nodes from selected outputs | hides graph mutation and conflicts with the PRD source-of-truth rules
Rejected: Put NewApiWG model-family payload rules in the composer | provider-specific schemas would leak into UI state
Confidence: high
Scope-risk: moderate
Directive: Keep provider payload normalization in provider/executor layers; do not move it into GenerationComposer
Tested: npm run test:run -- src/components/__tests__/GenerationComposer.test.tsx
Tested: npm run test:run -- src/components/__tests__/WorkflowCanvas.test.tsx
Tested: npm run test:run -- src/components/__tests__/GenerateVideoNode.test.tsx
Tested: npm run test:run -- src/app/api/generate/providers/__tests__/newapiwg.test.ts src/app/api/generate/__tests__/route.test.ts
Tested: npm run test:run -- src/store/execution/__tests__/videoProcessingExecutors.test.ts
Tested: npm run build
Not-tested: npm run lint fails because next lint is incompatible with the current Next 16 script configuration
The /api/generate route now submits Kie tasks and returns immediately
with polling metadata (taskId, pollProvider, etc.) instead of holding
the HTTP connection open for 2-10 minutes. Export buildMediaResponse
for reuse by the poll endpoint.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
maxDuration was 5 minutes which caused Next.js to kill the route handler
before Seedance 2.0 video generation polling could complete, resulting
in a network error on the client despite successful generation.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Instead of an inline text field, negative prompt is now a proper text
handle on the node that can receive connections from prompt/LLM nodes.
Updates the API route to merge dynamicInputs.negative_prompt into the
Veo parameters, and aligns the model schema definition accordingly.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Import generateWithGeminiVideo and route Veo model requests (modelId
starting with 'veo-') to the video provider before falling through
to the existing image generation path.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Nano Banana 2 (gemini-3.1-flash-image) supports a new searchTypes API
that allows enabling web search and image search independently. This adds
a useImageSearch toggle visible only for Nano Banana 2, while keeping the
existing Google Search toggle using the legacy format for Nano Banana Pro.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add generateAudio and videoFrameGrab cases to executeNode dispatcher
- Add selectedModel null guards in route.ts for all non-Gemini providers
- Fix ConnectionDropMenu wrap-around test (5->6 items after generateAudio)
- Fix audio duration edge cases (NaN guard in formatTime, truthiness check)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Resolve prompt from dynamicInputs.prompt in Gemini code path as fallback
- Return clear 400 error when prompt is provided but is not a string (corrupted data)
- Improve parallel batch error logging to include nodeId and nodeType for diagnosis
- Replace find() with for-loop to preserve index mapping between results and batch nodes
The helper returned string[] which didn't satisfy the ModelCapability[]
type in GenerationInput. Import and use the proper type.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Deduplicate the media response block (3d/video/audio/image) that was
copy-pasted across all 4 provider branches, and the mediaType-to-
capabilities ternary chain repeated 4 times.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add text-to-audio capability routing in all 4 provider blocks of
/api/generate, handle audio output type responses, and add audio
MIME type mappings and extraction logic to /api/save-generation.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: extract schema utilities from generate route
Extract INPUT_PATTERNS, InputMapping, ParameterTypeInfo, getParameterTypesFromSchema,
coerceParameterTypes, and getInputMappingFromSchema into schemaUtils.ts with 23 new tests.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: extract Gemini provider from generate route
Move MODEL_MAP and generateWithGemini() into providers/gemini.ts.
Remove unused imports (GoogleGenAI, uploadImageForUrl, shouldUseImageUrl,
deleteImages, ProviderModel) from route.ts.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: extract Replicate provider from generate route
Move generateWithReplicate() into providers/replicate.ts.
It imports schema utilities from schemaUtils.ts.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: extract fal.ai, Kie.ai, WaveSpeed providers from generate route
Move generateWithFalQueue, generateWithKie, generateWithWaveSpeed and all
their helpers into providers/fal.ts, providers/kie.ts, providers/wavespeed.ts.
Add barrel export at providers/index.ts.
route.ts reduced from 2,838 to 518 lines (POST handler + routing only).
Re-export clearFalInputMappingCache from route.ts for test backward compat.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: extract API header builder from workflow store
Replace 6 duplicated header-building if/else chains (across executeWorkflow
and regenerateNode) with buildGenerateHeaders() and buildLlmHeaders() utility
functions. Add 13 new tests.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: extract execution utilities from workflow store
Move groupNodesByLevel, chunk, revokeBlobUrl, clearNodeImageRefs,
and concurrency settings into executionUtils.ts with 21 new tests.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: extract getConnectedInputs and validateWorkflow as pure functions
Move getConnectedInputs (136 lines) and validateWorkflow (59 lines) into
connectedInputs.ts as pure functions taking (nodeId, nodes, edges).
Store methods become thin wrappers. Add 22 new tests.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: define node executor interface and context types
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: extract simple node executors
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: unify nanoBanana execution into shared executor
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: unify generateVideo execution into shared executor
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: unify llmGenerate execution into shared executor
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: unify splitGrid execution into shared executor
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: unify videoStitch and easeCurve executors
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: replace switch/if-else with node executor registry
Wire extracted executors into workflowStore, replacing the ~1000-line
switch statement in executeWorkflow and ~750-line if-else chain in
regenerateNode with executor registry calls. Store reduced from 3,402
to 1,782 lines.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: final cleanup of workflowStore module structure
Remove dead trackSaveGeneration function (no longer called after executor
extraction) and unused type imports. Store: 3,838 -> 1,721 lines (55% reduction).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: video regeneration spinner stuck + cancel error message
Fix video element not reloading on regeneration by adding key prop tied
to video history ID. Fix cancel showing timeout message by passing
"user-cancelled" reason through AbortController and checking it in all
three executor catch blocks.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: address PR #59 review — 18 bug, security, and robustness fixes
Providers:
- fal: don't cache failed schema fetches, remove duplicate param spread, SSRF-validate mediaUrl
- gemini: remove redundant header overrides, fix nano-banana model ID
- kie: module-level MAX_MEDIA_SIZE, post-download size check, upload size guard
- replicate: remove duplicate param spread, validate modelId format
- wavespeed: post-download size check, NaN-safe parseInt, prevent prompt overwrite
Executors:
- generateVideo: validate stored fallback inputs same as regular path
- nanoBanana: cap imageHistory at 50 entries
- splitGrid: handle img.onerror with warning + null data
- videoProcessing: add FileReader onerror/onabort handlers
Utilities:
- connectedInputs: use ?? instead of || for outputText/template
- executionUtils: guard orphan children in topological sort, validate chunk() size
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: additional PR review fixes — SSRF, data semantics, sync tracking, error propagation
Providers:
- replicate: add SSRF validation on mediaUrl before fetch
- fal/kie/replicate/wavespeed: large-video returns data="" with url set
instead of putting raw URL in data field (violates GenerationOutput.data semantic)
- kie: remove stale WaveSpeed JSDoc block from kie provider file
- route.ts: update all output handling to support data="" + url for large videos
Executors:
- generateVideoExecutor/nanoBananaExecutor: track save-generation fetch in
pendingImageSyncs via new trackSaveGeneration context method so auto-save
waits for server ID resolution
- videoProcessingExecutors: throw errors instead of returning in validation
branches (encoderSupported, insufficient videos, no video connected) so
Promise.allSettled sees rejections
- types.ts: add trackSaveGeneration to NodeExecutionContext interface
- workflowStore: wire trackSaveGeneration into both execution context sites
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: content-type fallback, replicate polling/validation, atomic gallery append
- fal.ts: Use model capabilities to determine fallback content-type
(video/mp4 vs image/png) instead of defaulting to video for all models
- replicate.ts: Increase polling timeout to 10 min for video models;
filter output array to valid strings before URL validation
- nanoBananaExecutor.ts: Replace read-modify-write on outputGallery with
atomic appendOutputGalleryImage using Zustand set() callback
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: extract _buildExecutionContext to deduplicate NodeExecutionContext construction
Replace identical inline context creation in executeWorkflow and regenerateNode
with a single shared helper, ensuring consistent behavior across both code paths.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: add keyboard shortcuts help dialog
- New KeyboardShortcutsDialog component showing all available shortcuts
- Accessible via ? button in header and ? key on canvas
- Groups shortcuts by category: General, Add Nodes, Layout, Canvas
- Auto-detects Mac vs Windows/Linux for modifier key display
- Follows existing modal patterns (Escape to close, backdrop click)
* feat: add keyboard shortcuts dialog and 3D GLB viewer node
- Keyboard shortcuts: press ? or click keyboard icon to view all shortcuts
- 3D GLB viewer: upload .GLB files, orbit/rotate, capture as image output
- Fix: restore trailing newline in gemini.ts provider
* fix: make Prompt node editable when connected to LLM node
When a Prompt node receives text from an LLM Generate node, the text
should be editable by the user. Previously, the textarea was marked as
readOnly when there was an incoming text connection, preventing edits
in both the main view and the expanded modal.
Changes:
- Track last received LLM output to detect when it changes
- Only update prompt when LLM node runs again (output changes)
- Allow user edits in between LLM runs
- Remove readOnly restriction on textarea
- Update placeholder text to indicate editability
This allows users to edit the LLM-generated text, which will be
replaced only when the LLM node runs again, not on every render.
Fixes#1470061054246518927
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
* fix: PR review fixes — lazy-load three.js, blob URL cleanup, missing integrations
- Remove unused `useLoader` import from GLBViewerNode
- Add useEffect cleanup to revoke blob URL on node unmount (memory leak)
- Replace alert() with useToast for file validation errors
- Lazy-load GLBViewerNode via next/dynamic to avoid bundling three.js for all users
- Add glbViewer to ConnectionDropMenu IMAGE_SOURCE_OPTIONS
- Add glbViewer case to quickstart validation createDefaultNodeData (build fix)
- Restore trailing newline in gemini.ts
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: address PR #64 review — blob leak, error handling, tests, docs
Move @types/three to devDependencies, fix blob URL cleanup to track
changes (not just unmount), add try/catch to capture and scene disposal,
remove GLBViewerNode from barrel export to preserve lazy-loading, fix
JSDoc, update CLAUDE.md with glbViewer node and ? shortcut, add
connectedInputs and nodeDefaults tests for glbViewer.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: improve GLB viewer — fix viewport sizing, add spot light, handle labels, scroll isolation
- Fix 3D viewport not filling node by adding resize={{ offsetSize: true }} for correct measurement in CSS-transformed React Flow nodes
- Replace grid/axis indicators with spot light for cleaner scene
- Add labeled input (3D) and output (Image) handles matching generate node style
- Add native wheel event listener to prevent graph zoom/pan while scrolling to zoom 3D model
- Remove redundant CSS workarounds (absolute positioning style, extra relative wrapper)
- Add edge-to-edge viewport via contentClassName (removes default padding when GLB loaded)
- Overlay controls bar on viewport with gradient background
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: add customizable canvas navigation settings (#65)
* feat: add customizable canvas navigation settings
Add a new settings modal that allows users to configure canvas navigation preferences:
- Pan Mode: Space + Drag (default), Middle Mouse, or Always On (ComfyUI-style)
- Zoom Mode: Alt + Scroll (default), Ctrl + Scroll, or Scroll (no modifier)
- Selection Mode: Click (default) or Alt + Drag
Features:
- Settings persist in localStorage
- New canvas settings button in header (monitor icon)
- Modal UI with radio button groups for each setting
- React Flow props dynamically configured based on user preferences
- Updated wheel event handler to respect zoom mode settings
This addresses user requests for ComfyUI/TouchDesigner-style navigation where
panning and zooming work without holding modifier keys.
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
* fix: PR #65 review — altDrag selection bug, icon, dedup, tests
- Fix altDrag selection mode: use selectionKeyCode="Alt" instead of
selectionOnDrag=true which made drag-select work without any key
- Replace misleading video camera icon with arrows-pointing-out for
canvas navigation settings button
- Extract duplicated settings buttons JSX into settingsButtons variable
shared by both configured/unconfigured project branches
- Add defensive useEffect to sync CanvasSettingsModal state on open
- Add missing semicolon in types/index.ts canvas re-export
- Add localStorage tests for getCanvasNavigationSettings and
saveCanvasNavigationSettings
- Add integration tests for updateCanvasNavigationSettings
- Fix WorkflowCanvas test mocks to include canvasNavigationSettings
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: move canvas navigation settings into Project Settings modal
Consolidate canvas settings from standalone CanvasSettingsModal into a
new "Canvas" tab in ProjectSetupModal. Remove "like ComfyUI" text, add
Shift+Drag selection mode, and fix save button to bottom of dialog.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: remove double semicolon in types/index.ts
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com>
* feat: add executeSelectedNodes with topo sort, abort, and downstream propagation
Implements the "Run selected nodes" feature (PR #61) with fixes for all 8
review issues: topological ordering via groupNodesByLevel, AbortController
cancellation, glbViewer case coverage, AbortError filtering, downstream
consumer propagation with dedup guard, and dynamic button label. Also
extracts saveLogSession helper to replace 4 inline duplicates across
executeWorkflow and regenerateNode.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: 3D model saving, error handling, and node improvements
- Save 3D models (GLB) to generations folder via save-generation API
- Add try-catch to WaveSpeed polling loop to handle network errors
- Skip downloading 3D model bodies in WaveSpeed (return URL directly)
- Fix schemaUtils false-positive matching for short property names
- Stabilize GenerateImageNode output handle ID (always "image")
- Fix GLBViewerNode: block-body forEach lint, store-based resize
- Use useRef instead of useState for PromptNode lastReceivedText
- Add executeGlbViewer tests and abort signal support
- Fix annotation stale pass-through when upstream image changes
- Skip error status on AbortError in all executor catch blocks
- Revoke glbUrl blob URLs in clearNodeImageRefs
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* docs: require per-task commits in multi-task plans
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: add Generate3DNodeData type definitions
Add "generate3d" to NodeType union, Generate3DNodeData interface,
Generate3DNodeDefaults, and update NodeDefaultsConfig.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: add generate3d node defaults and dimensions
Add default dimensions (300x300) and createDefaultNodeData case for
the generate3d node type.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: create generate3d executor
Extracted from nanoBananaExecutor's 3D handling code. Handles 3D model
generation via /api/generate with mediaType: "3d", cost tracking, and
auto-save to generations folder.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: create Generate3DNode component
Dedicated node component for 3D model generation, modeled after
GenerateVideoNode. Features: dynamic input handles from schema,
3D output handle, model browsing, parameter controls, status overlays.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: register generate3d node type across codebase
Register in execution index, workflowStore (executeWorkflow,
regenerateNode, executeSelectedNodes), nodes index, WorkflowCanvas
(nodeTypes, getNodeHandles, minimap color, findCompatibleHandle).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: add generate3d source output in connectedInputs
Add generate3d case in getSourceOutput() to return 3D URL data,
enabling downstream nodes to receive 3D model output.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: route 3D models to generate3d in ModelSearchDialog
Add "3d" capability filter, update handleSelectModel to detect 3D
models and create generate3d nodes with capabilities passed through,
update recent models filter for 3D.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: update ConnectionDropMenu 3D source to use generate3d
Change THREE_D_SOURCE_OPTIONS from nanoBanana to generate3d with
cube icon and "Generate 3D" label.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: remove 3D code from GenerateImageNode and nanoBananaExecutor
- Remove text-to-3d/image-to-3d from IMAGE_CAPABILITIES
- Remove is3D memo and conditional handle/label styling
- Remove 3D output preview block
- Remove output3dUrl from model change handlers
- Remove is3DModel detection and mediaType from nanoBanana executor
- Remove 3D response handling from nanoBanana executor
- Remove output3dUrl from NanoBananaNodeData and its defaults
- Remove 3D fallback from nanoBanana in connectedInputs
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: update tests and remaining registrations for generate3d
- Add Generate3DNode.test.tsx with render, handle, and output tests
- Add 3D handle tests to ConnectionDropMenu.test.tsx
- Add 3D model routing test to ModelSearchDialog.test.tsx
- Update ModelSearchDialog test expectations for capabilities passthrough
- Register generate3d in executeNode.ts dispatcher
- Add generate3d to WorkflowCanvas keyboard shortcut dimensions
- Add generate3d to quickstart validation (VALID_NODE_TYPES, dimensions, defaults)
- Add generate3d to chat tools (VALID_NODE_TYPES, description)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: unwrap array inputs for non-array API parameters
When multiple images are connected to a generation node, dynamicInputs
aggregates them into arrays. Providers wrapped single values into arrays
for array-typed params, but never unwrapped arrays for string-typed params,
causing API failures (e.g. fal's image_url receiving an array instead of
a string). Add symmetric unwrap logic across all four providers.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: move 3D from floating action bar to Generate submenu
Replace top-level glbViewer button with generate3d option in the
Generate combo dropdown, between Video and Text (LLM).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: hide CostIndicator when non-Gemini providers are in workflow
The pricing calculator only has reliable data for Gemini models. When
non-Gemini providers (fal, replicate, kie, wavespeed) are used, the
displayed cost is incomplete/misleading. Add hasNonGeminiProviders()
utility and use it to hide the CostIndicator entirely in that case.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: add 3D capability badges to ModelSearchDialog
Add text-to-3d and image-to-3d capability badge styling.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: add play/run icon to LLM Generate node title bar
Wire onRun and isExecuting props to BaseNode, matching the pattern
used in GenerateImageNode and GenerateVideoNode.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: clear imageRef when user provides new image to prevent stale saves
When a user pastes/uploads a new image into a node that already has a
saved imageRef, the externalization logic incorrectly assumes the base64
is just hydrated data and discards it. Clear *Ref fields whenever new
image data is set so the save logic correctly persists the new image.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: guard against undefined from empty array unwrap in fal provider
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: add 3x2 grid layout option to split grid node
Change layout selector from count-based to layout-based, allowing both
2x3 and 3x2 (portrait) layouts that produce 6 images. Now shows 7
layout options (2x2, 1x5, 2x3, 3x2, 2x4, 3x3, 2x5) with RxC labels.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: charlieshoelace <charlieshoelace@gmail.com>
- Fix 3D viewport not filling node by adding resize={{ offsetSize: true }} for correct measurement in CSS-transformed React Flow nodes
- Replace grid/axis indicators with spot light for cleaner scene
- Add labeled input (3D) and output (Image) handles matching generate node style
- Add native wheel event listener to prevent graph zoom/pan while scrolling to zoom 3D model
- Remove redundant CSS workarounds (absolute positioning style, extra relative wrapper)
- Add edge-to-edge viewport via contentClassName (removes default padding when GLB loaded)
- Overlay controls bar on viewport with gradient background
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Providers:
- replicate: add SSRF validation on mediaUrl before fetch
- fal/kie/replicate/wavespeed: large-video returns data="" with url set
instead of putting raw URL in data field (violates GenerationOutput.data semantic)
- kie: remove stale WaveSpeed JSDoc block from kie provider file
- route.ts: update all output handling to support data="" + url for large videos
Executors:
- generateVideoExecutor/nanoBananaExecutor: track save-generation fetch in
pendingImageSyncs via new trackSaveGeneration context method so auto-save
waits for server ID resolution
- videoProcessingExecutors: throw errors instead of returning in validation
branches (encoderSupported, insufficient videos, no video connected) so
Promise.allSettled sees rejections
- types.ts: add trackSaveGeneration to NodeExecutionContext interface
- workflowStore: wire trackSaveGeneration into both execution context sites
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Move generateWithFalQueue, generateWithKie, generateWithWaveSpeed and all
their helpers into providers/fal.ts, providers/kie.ts, providers/wavespeed.ts.
Add barrel export at providers/index.ts.
route.ts reduced from 2,838 to 518 lines (POST handler + routing only).
Re-export clearFalInputMappingCache from route.ts for test backward compat.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Move generateWithReplicate() into providers/replicate.ts.
It imports schema utilities from schemaUtils.ts.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Move MODEL_MAP and generateWithGemini() into providers/gemini.ts.
Remove unused imports (GoogleGenAI, uploadImageForUrl, shouldUseImageUrl,
deleteImages, ProviderModel) from route.ts.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Extract INPUT_PATTERNS, InputMapping, ParameterTypeInfo, getParameterTypesFromSchema,
coerceParameterTypes, and getInputMappingFromSchema into schemaUtils.ts with 23 new tests.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Validate uploadUrl and fileUrl from fal CDN initiate response through
validateMediaUrl() and HTTPS check before performing the PUT request.
Throws if either URL is missing or fails validation.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Validate fal-provided status_url and response_url against validateMediaUrl()
and domain allowlist (queue.fal.run) before using them for polling. Falls back
to constructed URLs on validation failure.
Also updates mockFalCdnUpload test helper to match the two-step CDN upload
flow (initiate + PUT) introduced in 3fea13e.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
WaveSpeed's CloudFront CDN returns `binary/octet-stream` which wasn't
caught by the old `!== "application/octet-stream"` check, producing
unplayable data URLs. Now only accepts content types starting with
`video/` or `image/`, falling back to `video/mp4` or `image/png`.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The old single-POST endpoint (fal.ai/api/storage/upload) returns 404.
Switched to the current fal.ai upload API: initiate via
rest.alpha.fal.ai/storage/upload/initiate, then PUT binary data to
the returned signed URL.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Delete unused generateWithFal function (replaced by generateWithFalQueue)
- Add 20MB size guard in uploadImageToFal to prevent memory spikes
- Add refresh button to ModelSearchDialog that clears model and schema
localStorage caches plus in-memory fetch cache
- Wire "Try Again" error button to also clear stale caches
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace blocking fal.run endpoint with async queue.fal.run + 1s polling
to match Replicate/WaveSpeed behavior. Upload base64 images to fal CDN
before queue submission to avoid large inline payloads. Cache schema
mappings in-memory (30min TTL) to eliminate redundant API calls per
generation. Also fix localStorage cache leaking between tests.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Derive provider capabilities from mediaType instead of hardcoding
text-to-image, so isVideoModel is correct for video generations.
Treat application/octet-stream as unhelpful content-type and fall back
to video/mp4 or image/png based on model type. Add safety net in
save-generation to override .bin extension when media type is known.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add Veo 3 and Veo 3 Fast models (text-to-video + image-to-video) with
custom API integration using Kie's /api/v1/veo/ endpoints and numeric
successFlag polling. Also document Kie model integration SOP in CLAUDE.md.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Refactor executeSingleNode error paths to throw instead of setting global
isRunning/currentNodeIds state, which corrupted sibling nodes running in
parallel via Promise.allSettled
- Validate WaveSpeed providedPollUrl with SSRF check before use
- Sanitize WaveSpeed taskId in provider with path-traversal validation
- Revoke blob URLs (URL.revokeObjectURL) before replacing outputVideo to
prevent memory leaks from large video files
- Move pause edge cleanup to outer executeWorkflow loop
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add validateMediaUrl() utility to block private IPs, localhost, and
non-HTTP protocols before server-side fetches (with tests)
- URL-encode Kie taskId in poll URL to prevent injection
- Validate WaveSpeed modelId against path traversal in both route.ts
and wavespeed.ts provider
- Add 500MB Content-Length check before downloading media at all 3
fetch sites (Kie, WaveSpeed route, WaveSpeed provider)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Update stale test mocks from currentNodeId to currentNodeIds in
AudioInputNode, EaseCurveNode, and VideoStitchNode tests. Replace
hard 401 return for missing FAL API key with console.warn, allowing
requests to proceed without auth (rate-limited) since
generateWithFal() already handles null keys gracefully.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add nano-banana-pro model to Kie provider with schema and image input key
- Reorder Kie generation to process dynamicInputs before fallback image
uploads, preventing double-uploads when schema-mapped connections exist
- Update test mocks for deduplicatedFetch, useProviderApiKeys, and static
handle changes across GenerateImageNode, GenerateVideoNode,
ModelParameters, and ModelSearchDialog tests
- Add test commands to CLAUDE.md
- Fix SplitGridSettingsModal grid preview count (5 → 6 options)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Remove 36 unsupported Kie.ai models (14 image, 22 video) and add 2 new
ones (kling-2.6/motion-control, topaz/video-upscale), bringing the total
to 22 models. Fix parameter schemas to match official Kie.ai docs:
seedream quality param, flux-2 input_urls, grok-imagine video mode/duration,
kling sound param, wan resolution param, and singular vs array image keys.
Remove dead Veo/Kontext/Runway helper functions and endpoint branching —
all remaining models use the standard createTask endpoint.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add ~38 new image/video models (Imagen 4, FLUX.2, Flux Kontext,
Grok Imagine, Ideogram, Qwen, Runway, Hailuo, Seedance variants,
Kling older, Wan older) with schemas and endpoint routing for
dedicated APIs (Flux Kontext, Runway). Add pageUrl to all Kie
models and update ModelSearchDialog to use it.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Added validation to ensure fal.ai API key is provided in requests, returning a 401 status with an appropriate error message if not configured.
- Updated related tests to reflect the new API key requirement and adjusted assertions for expected responses.
- Modified documentation to clarify that fal.ai API key is now mandatory for model access.
GPT Image 1.5 only supports 'aspect_ratio', not 'size'.
Remove stale 'size' values from old workflow data before sending.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
weige
TianYun