Changelog

Track all updates to FloImg packages, Studio, and Cloud platform.

v0.21.12

SDK
View on GitHub

Docker

  • ~ Skip cache for runner stage to ensure build args are embedded

v0.21.11

SDK
View on GitHub

Docker

  • ~ Inherit build args in production stage for deployment visibility

v0.21.8

SDK
View on GitHub

@teamflojo/floimg 0.17.0

  • + Consolidated validation with semantic checks for end-to-end consistency

@teamflojo/floimg-mcp 0.2.0

  • + Add pre-execution validation to run_pipeline tool

FloImg Studio Backend

  • + AI workflow generator retry loop for invalid workflows
  • + Validation error mapping for UI highlighting

v0.21.2

SDK
View on GitHub

Build Fixes

  • ~ Include floimg-vectorize COPY instructions in Dockerfile

v0.21.1

SDK
View on GitHub

Build Fixes

  • ~ Add floimg-vectorize to npm publish workflow
  • ~ Add floimg-vectorize to Docker build steps

v0.20.3

SDK
View on GitHub

FloImg Studio Backend

  • ~ Add composite transform documentation to AI workflow generator

v0.20.1

Studio
View on GitHub

@teamflojo/floimg-studio-shared 0.9.1

  • ~ Accept "base" targetHandle for composite transform base image input
  • ~ Collect overlay images via "overlays[N]" handles with index preservation
  • ~ Support sparse overlay indices (e.g., [2, 5] instead of [0, 1])

@teamflojo/floimg-studio-backend 0.5.2

  • ~ Resolve and inject overlay images for composite transforms
  • ~ Validate missing overlays with clear error message instead of cryptic Sharp error

v0.20.0

Studio
View on GitHub

@teamflojo/floimg-studio-ui 0.9.0

  • + Execution Results Modal for viewing run details
  • + History indicator badge on History tab
  • + Per-node execution timing display
  • + Compare mode for execution history
  • + Pin favorite runs
  • + Export history to ZIP
  • - deps: Add jszip for client-side ZIP generation

v0.19.0

Studio
View on GitHub

@teamflojo/floimg-studio-ui 0.8.0

  • - **BREAKING**: "Images" tab renamed to "History" tab
  • - refactor: Remove "Generated images bar" from Toolbar
  • + Add execution history tracking with data URLs
  • + New `ExecutionHistory` component showing workflow run history
  • + Export `ExecutionRun`, `ExecutionRunOutput`, `SavedWorkflow` types

v0.18.0

Studio
View on GitHub

@teamflojo/floimg-studio-ui 0.7.0

  • - **BREAKING**: Removed `@teamflojo/floimg-templates` package dependency
  • - **BREAKING**: Removed template re-exports (`templates`, `getCategories`, `getTemplateById`, `getTemplatesByCategory`, `searchTemplates`, `resolveTemplate`)
  • + `TemplateGallery` now fetches templates from API with offline fallback
  • + Templates cached in localStorage for offline access after first fetch
  • + Bundled seed templates for air-gapped/offline cold starts
  • + Added `apiUrl` prop to `TemplateGallery` for custom API endpoints

@teamflojo/floimg-studio-shared 0.9.0

  • + Added `Template` and `TemplateCategory` types (migrated from floimg-templates)
  • - **REMOVED**: This package has been deleted. Templates are now served from the API.
  • - Migration: Use `fetch('https://api.floimg.com/api/templates')` to get templates
  • - Migration: Import `Template` type from `@teamflojo/floimg-studio-shared` instead

v0.17.0

Studio
View on GitHub

@teamflojo/floimg-studio-ui 0.6.0

  • + add `parentWorkflowId` state tracking for remix attribution chains
  • + add `loadRemixImage(imageUrl)` store method for creating Input nodes from remote URLs
  • + add `loadFromWorkflowJson(json, metadata)` store method for loading forked workflows with ID remapping
  • + handle `?remixImage=<url>` URL parameter for external remix links
  • + add `remixInfoSlot` prop to Toolbar for injecting attribution banners

@teamflojo/floimg-studio-shared 0.8.0

  • + add `imageUrl` field to `InputNodeData` for remote image URLs in remix flows

v0.15.7

SDK
View on GitHub

@teamflojo/floimg-studio-shared 0.7.0

  • + add `nodesToPipeline()` function for converting visual workflows to Pipeline format
  • + add `Pipeline`, `PipelineStep`, and `PipelineConversionResult` types
  • + 21 interface consistency tests for nodesToPipeline conversion
  • - refactor: unified step identifiers - node IDs used as variable names throughout

@teamflojo/floimg-mcp 0.1.1

  • - refactor: use canonical Pipeline format internally for execution

v0.15.4

Studio
View on GitHub

@teamflojo/floimg-studio-ui 0.5.3

  • ~ dark mode text visibility in NodePalette (description text was black/invisible)
  • ~ category label contrast in dark mode (zinc-400 for readability)
  • ~ remove colored left borders (was creating visual noise with 7 accent colors)
  • ~ make title text neutral (removed accent-colored titles)
  • + add grip dots for drag affordance (persistent visual cue that items are draggable)
  • ~ improve typography (description text 13px minimum)
  • ~ minimap dark mode (respects system color scheme)

v0.15.2

Studio
View on GitHub

@teamflojo/floimg-studio-ui 0.5.1

  • + export keyboard shortcuts components and hooks for library consumers
  • ~ ensure CSS theme is bundled in library build
  • * add verify-lib script to validate library build output

v0.14.0

SDK
View on GitHub

@teamflojo/floimg 0.12.0

  • + add `UsageEvent` and `UsageHooks` interfaces for AI cost observability

@teamflojo/floimg-openai 0.3.0

  • + add optional `hooks` config for usage event emission

@teamflojo/floimg-stability 0.3.0

  • + add optional `hooks` config for usage event emission

@teamflojo/floimg-google 0.6.0

  • + add optional `hooks` config for usage event emission

@teamflojo/floimg-replicate 0.2.0

  • + add optional `hooks` config for usage event emission

@teamflojo/floimg-xai 0.2.0

  • + add optional `hooks` config for usage event emission

@teamflojo/floimg-studio-backend 0.4.0

  • + collect and return usage events from workflow execution

v0.13.0

SDK
View on GitHub

@teamflojo/floimg-google 0.5.0

  • + add `enhancePrompt` option for Gemini generate and edit nodes
  • + export prompt enhancement utilities: `enhancePrompt()`, `detectPromptType()`, `isPromptDetailed()`
  • * improve TypeScript types with proper SDK type imports

@teamflojo/floimg-studio-ui 0.5.0

  • + comprehensive visual identity refresh with custom theming system
  • + gradient backgrounds for toolbar and sidebars
  • + color-coded palette items by node category
  • + node execution state animations (running shimmer, completed/error indicators)
  • + hide right sidebar when no node selected for cleaner canvas view
  • ~ rename cloud save provider from "floimg-cloud" to "cloud" for clarity

v0.12.0

SDK
View on GitHub

@teamflojo/floimg 0.11.0

  • + add fan-out, collect, and router pipeline primitives for iterative workflows
  • + add YAML example for iterative workflow pattern (`examples/pipeline-iterative.yaml`)
  • ~ improve error messages for router selection property validation
  • ~ add warnings when fan-out count/array length doesn't match output array length

@teamflojo/floimg-studio-ui 0.5.0

  • + add Fan-Out node for parallel branch execution
  • + add Collect node for gathering branch outputs
  • + add Router node for AI-driven selection from candidates
  • + add iterative workflow support to executor with SDK delegation
  • + update node palette with Iterative Workflow section

@teamflojo/floimg-studio-shared 0.5.0

  • + add FanOutNodeData, CollectNodeData, RouterNodeData types
  • + add iterative workflow node type definitions

@teamflojo/floimg-studio-backend 0.3.0

  • + add executor support for fan-out, collect, and router steps
  • + delegate branching execution to SDK client.run()

v0.11.0

Studio
View on GitHub

@teamflojo/floimg-studio-ui 0.4.0

  • + real-time workflow execution progress via SSE streaming
  • + output inspector modal for viewing full text/vision node responses
  • + AI generation cancellation support
  • + per-node execution status indicators (running → completed/error/skipped)
  • ~ add size/depth safeguards to JSON tree rendering (max depth 20, max array 500)

@teamflojo/floimg-studio-shared 0.4.0

  • + add SSE event types for execution and generation streaming
  • + add "skipped" status to ExecutionStepResult with skipReason field

@teamflojo/floimg-studio-backend 0.2.0

  • + add `/api/execute/stream` SSE endpoint for real-time execution progress
  • + add `/api/generate/workflow/stream` SSE endpoint with generation phases
  • + sequential execution with inline moderation and real-time callbacks
  • + partial failure handling - skip dependent steps when upstream fails

v0.10.0

SDK
View on GitHub

@teamflojo/floimg 0.9.0

  • - generator: shapes
  • - generator: shapes
  • + add 4 new shapes - `ellipse`, `triangle`, `polygon` (with `sides`), `star` (with `points`, `innerRadius`)
  • + add stroke support - `strokeColor`, `strokeWidth`
  • + add rotation parameter (0-360 degrees)
  • + add radial gradient type (`gradientType: 'radial'`)
  • + add checkerboard pattern (`patternType: 'checkerboard'`)

@teamflojo/floimg-studio-ui 0.3.4

  • + add conditional field visibility for shapes generator (only show relevant params)

v0.9.6

SDK
View on GitHub

All Plugin Packages

  • ~ update peer dependency on `@teamflojo/floimg` from `^0.2.0` to `^0.8.0`
  • + Updated Packages

v0.9.4

SDK
View on GitHub

@teamflojo/floimg-templates 0.2.0

  • + enhance all 12 single-node templates to multi-node workflows (4 nodes each)
  • + add 6 new multi-node workflow templates
  • + add `nodeCount` field to Template type
  • + add `getNodeCount()`, `getPipelineTemplates()`, `getTemplatesByComplexity()` helpers
  • * update template descriptions with clear JTBDs

v0.9.3

Repository
View on GitHub

Repository

  • ~ add floimg-templates to Docker build

v0.9.2

SDK
View on GitHub

@teamflojo/floimg-templates 0.1.0

  • ~ actually publish to npm (was missing from release workflow in v0.9.1)
  • # improve package documentation clarity

@teamflojo/floimg-studio-ui 0.3.2

  • ~ use workspace dependency for floimg-studio-shared (was pulling outdated npm version)

Repository

  • * add floimg-templates to release workflow
  • # improve templates package documentation

v0.9.1

SDK
View on GitHub

@teamflojo/floimg-templates 0.1.0

  • + new shared package for workflow templates
  • + templates organized by category (AI Workflows, Data Viz, Marketing, Utilities)
  • + `coreTemplates` (bundled, works offline) vs `allTemplates` (includes cloud-only templates)
  • + `resolveTemplate()` for legacy ID mapping
  • + `getStudioUrl()` helper for template deep links

@teamflojo/floimg-studio-ui 0.3.1

  • - refactor: use `@teamflojo/floimg-templates` as single source of truth
  • + template loading uses `resolveTemplate()` for backwards compatibility
  • * remove local template definitions (now in shared package)

v0.9.0

SDK
View on GitHub

@teamflojo/floimg-studio-ui 0.3.0

  • + context-aware AI generation error UI
  • + show "View Setup Guide" link for self-hosted deployments
  • + show "View Plans" link for tier limit errors (cloud deployments)
  • + show "Contact Support" link for service errors (paid users only)

@teamflojo/floimg-studio-shared 0.3.0

  • + add `GenerateStatusResponse` and `GenerateStatusReason` types
  • + add `reason`, `isCloudDeployment`, and `supportUrl` fields to status response
  • + add `reason` and `isCloudDeployment` fields to `/api/generate/status`
  • + server-side logging when GOOGLE_AI_API_KEY is not configured
  • ! use `GOOGLE_AI_API_KEY` instead of `GEMINI_API_KEY` for workflow generation

@teamflojo/floimg-google 0.4.0

  • ! remove `GEMINI_API_KEY` fallback, use `GOOGLE_AI_API_KEY` only

@teamflojo/floimg 0.8.0

  • ~ update MCP server error messages to reference `GOOGLE_AI_API_KEY`

Breaking Changes

  • - **GEMINI_API_KEY no longer supported** - Use `GOOGLE_AI_API_KEY` instead for all Google AI features including AI Workflow Generator

v0.8.0

SDK
View on GitHub

@teamflojo/floimg-studio-ui 0.2.0

  • + AI-powered workflow generation with Gemini 3 Pro
  • + add node duplication with Cmd+D
  • + add prompt input handle to gemini-generate node
  • + multi-reference image support for Gemini nodes
  • + support generator/transform outputs as reference images
  • + enhance AI workflow generator with structured output examples
  • + add advanced workflow examples for AI generation
  • ~ use screenToFlowPosition for correct node drop coordinates
  • ~ set outputSchema on text nodes when loading AI workflows
  • ~ transform Gemini parametersJson to parameters object
  • ~ distinguish cloud save from filesystem save in inspector

@teamflojo/floimg-google 0.3.1

  • + support GEMINI_API_KEY env var as fallback

v0.7.1

SDK
View on GitHub

Fixed

  • - Added `@teamflojo/floimg-xai` to npm publish workflow
  • - Added floimg-xai to Docker build for FloImg Studio

v0.7.0

SDK
View on GitHub

Added

  • + New Plugin: xAI Grok
  • - **`@teamflojo/floimg-xai`** - xAI integration (NEW)
  • + CLI AI Commands
  • - **`floimg ai text`** - Generate text using configured AI providers
  • - **`floimg ai vision`** - Analyze images using configured AI vision models
  • - **MCP provider discovery** - Dynamic provider detection for MCP server
  • + Google AI Enhancements
  • - **Gemini Generate** - Text-to-image generation with Gemini
  • - **Gemini Edit** - AI-powered image editing transform with dynamic prePrompt
  • - **New parameters**: `aspectRatio`, `imageSize`, `groundingWithSearch`
  • - **Updated models**: Gemini 2.5-flash and 3-flash-preview
  • + FloImg Studio
  • - **AI text/vision nodes** - Multi-output support for AI transforms
  • - **Dynamic text inputs** - Text input support for AI transforms
  • - **Provider routing** - AI transform provider routing and API key injection
  • - **providerLabel** - Cleaner display names for AI nodes

v0.6.0

SDK
View on GitHub

Added

  • + Fluent API
  • - **FluentBuilder class**: Chainable syntax for building image workflows
  • - **createFluent() factory**: Create fluent facade for custom client configurations
  • - **floimg singleton**: Pre-configured fluent API export for quick usage
  • + AI Provider Packages
  • - **`@teamflojo/floimg-openai`** - OpenAI integration (extracted from core)
  • - **`@teamflojo/floimg-stability`** - Stability AI integration
  • - **`@teamflojo/floimg-google`** - Google AI integration
  • - **`@teamflojo/floimg-replicate`** - Replicate integration (NEW)
  • - **`@teamflojo/floimg-ollama`** - Ollama local AI integration
  • + Transform Provider Architecture
  • - **Self-dispatching providers**: TransformProvider now handles its own operation routing
  • - **I/O type metadata**: Schemas include inputType/outputType for visual builder validation
  • - **AI transforms**: Support for AI-powered image transformations
  • - **Four transform providers**: sharp (built-in), stability-transform, openai-transform, replicate-transform

Changed

  • - **Studio executor refactored**: Now uses floimg's core pipeline runner instead of custom wave execution
  • - **Pipeline runner**: Vision and text step kinds now properly tracked in dependency graph
  • - Extracted OpenAI provider to separate `@teamflojo/floimg-openai` package
  • - TransformProvider interface now requires `transform()` method
  • - Class renamed from `Floimg` to `FloImg` for consistency

v0.5.0

SDK
View on GitHub

Added

  • + FloImg Studio
  • - **Visual workflow builder**: Drag-and-drop interface for creating image pipelines
  • - **Node-based editor**: React Flow canvas with 20+ node types
  • - **Template gallery**: Pre-built workflows for common tasks
  • - **Real-time preview**: See results as you build
  • + npm Packages
  • - `@teamflojo/floimg-studio-ui` - React components for building custom studio UIs
  • - `@teamflojo/floimg-studio-shared` - Shared TypeScript types
  • + Docker Support
  • - Pre-built Docker images on GitHub Container Registry
  • - `docker run ghcr.io/flojoinc/floimg-studio` for instant deployment
  • + CI/CD
  • - GitHub Actions workflow for npm publishing
  • - Docker image builds on release

Changed

  • - Migrated FloImg Studio into monorepo (previously separate repo)
  • - Updated documentation for self-hosting options

v0.2.0

SDK
View on GitHub

Added

  • + AI Vision & Text Generation
  • - **Vision analysis**: `analyzeImage()` method for AI-powered image understanding
  • - **Text generation**: `generateText()` method for LLM text completion
  • - **DataBlob type**: New output type for text/JSON results from AI operations
  • - **VisionProvider interface**: Standardized interface for vision AI providers
  • - **TextProvider interface**: Standardized interface for text AI providers
  • + AI Providers
  • - **OpenAI**: GPT-4 Vision for image analysis, GPT-4 for text generation
  • - **Anthropic**: Claude vision and text support (via config)
  • - **Google Gemini**: Gemini vision and text support (via config)
  • - **Ollama**: Local AI models (LLaVA for vision, Llama for text)
  • - **LM Studio**: Local model support via OpenAI-compatible API
  • + New Package
  • - `@teamflojo/floimg-ollama`: Plugin for local AI models via Ollama
  • + MCP Server
  • - `analyze_image` tool: AI vision analysis in MCP workflows
  • - `generate_text` tool: AI text generation in MCP workflows

Changed

  • - Extended `getCapabilities()` to include AI provider information
  • - Updated config types for multi-provider AI support

Notes

  • - All AI features are optional - FloImg works without any AI configured
  • - AI providers require respective API keys or local model setup

v0.1.0

SDK
View on GitHub

Added

  • + Core Library
  • - **Three core operations**: generate, transform, save
  • - **Multiple interfaces**: JavaScript API, CLI, YAML pipelines, MCP server
  • - **Capability discovery**: Runtime schema inspection via `getCapabilities()`
  • - **Parallel pipeline execution**: Automatic dependency graph analysis and wave-based execution
  • + Built-in Generators
  • - `shapes` - SVG gradients, circles, rectangles, patterns
  • - `openai` - DALL-E image generation
  • + Transform Operations
  • - Format conversion (SVG to PNG/JPEG/WebP/AVIF)
  • - Resize with fit modes
  • - Composite (layer images)
  • - Filters: blur, sharpen, grayscale, negate, normalize, threshold, modulate, tint
  • - Borders: extend, extract, roundCorners
  • - Text: addText, addCaption
  • - 8 preset filters: vintage, vibrant, blackAndWhite, dramatic, soft, cool, warm, highContrast
  • + Save Providers
  • - Filesystem (default, zero-config)
  • - S3-compatible storage (AWS S3, Tigris, Cloudflare R2)
  • - Smart URI routing: `./local/path` vs `s3://bucket/key`
  • + CLI
  • - `floimg generate` - Generate images
  • - `floimg transform` - Transform images
  • - `floimg save` - Save to filesystem or cloud
  • - `floimg run` - Execute YAML pipelines
  • - `floimg doctor` - Check configuration
  • - `floimg mcp install` - Set up MCP integration
  • + MCP Server
  • - Session workspace for efficient multi-step workflows
  • - Image ID tracking (no byte passing between operations)
  • - `generate_image`, `transform_image`, `save_image`, `run_pipeline` tools
  • + Plugins
  • - `@teamflojo/floimg-quickchart` - Chart.js charts via QuickChart API
  • - `@teamflojo/floimg-d3` - D3 data visualizations
  • - `@teamflojo/floimg-mermaid` - Mermaid diagrams
  • - `@teamflojo/floimg-qr` - QR code generation
  • - `@teamflojo/floimg-screenshot` - Website screenshots via Playwright

Dependencies

  • - `sharp` - Image processing
  • - `@resvg/resvg-js` - SVG rendering
  • - `@napi-rs/canvas` - Text rendering
  • - `@aws-sdk/client-s3` - S3 uploads
  • - `openai` - DALL-E integration
  • - `@modelcontextprotocol/sdk` - MCP server
  • - **Major** (x.0.0): Breaking changes to API
  • - **Minor** (0.x.0): New features, backward compatible
  • - **Patch** (0.0.x): Bug fixes, backward compatible
  • - [GitHub Repository](https://github.com/FlojoInc/floimg)
  • - [npm Package](https://www.npmjs.com/package/@teamflojo/floimg)
  • - [Documentation](https://github.com/FlojoInc/floimg#readme)
GitHub Releases npm Packages