{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"github_mcp-google-labs-code-stitch-skills","slug":"mcp-google-labs-code-stitch-skills","name":"stitch-skills","type":"mcp","url":"https://github.com/google-labs-code/stitch-skills","page_url":"https://unfragile.ai/mcp-google-labs-code-stitch-skills","categories":["mcp-servers","app-builders"],"tags":[],"pricing":{"model":"open_source","free":true,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"github_mcp-google-labs-code-stitch-skills__cap_0","uri":"capability://tool.use.integration.agent.agnostic.skill.installation.and.discovery","name":"agent-agnostic skill installation and discovery","description":"Automatically detects active AI coding agents (Antigravity, Gemini CLI, Claude Code, Cursor) on the developer's system and installs standardized skills into agent-specific directories without manual configuration. Uses a skills CLI that scans the filesystem for agent installation paths and deploys skills following the Agent Skills open standard directory structure, enabling write-once-run-anywhere skill distribution across heterogeneous agent platforms.","intents":["I want to install skills once and have them work across multiple AI coding agents without reconfiguring each agent","I need to distribute skills to teams using different coding agents without maintaining agent-specific versions","I want to discover what skills are available and install them with a single command"],"best_for":["teams using multiple AI coding agents (Cursor, Claude Code, Gemini CLI)","skill developers building for cross-agent compatibility","enterprises standardizing on agent-agnostic skill ecosystems"],"limitations":["requires agents to be installed and discoverable via standard filesystem paths","no support for agents with custom installation directories or containerized deployments","agent detection is filesystem-based, not API-based, so may fail with non-standard setups"],"requires":["Node.js 18+","npm 7+","at least one supported AI coding agent installed (Antigravity, Gemini CLI, Claude Code, or Cursor)","Stitch MCP Server (latest) for most skills to function"],"input_types":["filesystem paths","agent configuration metadata"],"output_types":["installed skill directories","agent-specific configuration files"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"github_mcp-google-labs-code-stitch-skills__cap_1","uri":"capability://tool.use.integration.standardized.skill.instruction.and.execution.framework","name":"standardized skill instruction and execution framework","description":"Provides a structured directory convention (SKILL.md, scripts/, resources/, examples/) that enables AI agents to consistently discover task instructions, validate outputs, and learn from reference implementations. Each skill follows the Agent Skills open standard, allowing agents to parse SKILL.md for mission/workflow/success criteria, execute validation scripts for quality enforcement, and reference example outputs for in-context learning without agent-specific adaptation.","intents":["I want to create a skill that any AI agent can understand and execute without custom integration code","I need agents to validate their own outputs against success criteria defined in the skill","I want to provide reference examples so agents can learn the expected output format"],"best_for":["skill developers building for multiple agent platforms","teams establishing consistent skill quality standards","organizations standardizing on open skill ecosystems"],"limitations":["agents must implement parsing logic for SKILL.md format; no guarantee of consistent interpretation across agent implementations","validation scripts are agent-agnostic but may require agent-specific wrappers for execution","no built-in versioning or backward compatibility mechanism for skill format evolution"],"requires":["adherence to Agent Skills open standard directory structure","SKILL.md file with task description, step-by-step procedures, and success criteria","scripts/ directory with executable validation/integration programs","resources/ directory with reference materials","examples/ directory with syntactically valid reference implementations"],"input_types":["markdown documentation (SKILL.md)","executable scripts (bash, node, python)","reference code examples"],"output_types":["structured skill metadata","validation results","agent-executable task definitions"],"categories":["tool-use-integration","planning-reasoning"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"github_mcp-google-labs-code-stitch-skills__cap_10","uri":"capability://memory.knowledge.reference.implementation.learning.and.in.context.examples","name":"reference implementation learning and in-context examples","description":"Provides syntactically valid reference implementations in the examples/ directory of each skill, enabling agents to learn expected output formats, coding patterns, and best practices through concrete examples. Agents can reference these examples during code generation to understand the desired output structure, style, and quality level, improving generation accuracy through in-context learning without requiring explicit instruction in SKILL.md.","intents":["I want agents to learn the expected output format by examining reference examples","I need to demonstrate coding patterns and best practices through concrete examples","I want to improve code generation quality by providing agents with high-quality reference implementations"],"best_for":["skill developers wanting to guide agent behavior through examples","teams establishing coding standards and patterns for generated code","organizations building skills with complex output requirements"],"limitations":["examples must be manually created and maintained; no automatic example generation","agents may not effectively learn from examples if they're too complex or too simple","examples are static; no mechanism to update examples when best practices evolve","no guarantee that agents will follow example patterns; depends on agent implementation"],"requires":["syntactically valid reference implementations in examples/ directory","examples covering common use cases and edge cases","examples demonstrating best practices and coding standards"],"input_types":["reference implementation files","example descriptions and annotations","design specifications that examples implement"],"output_types":["agent-generated code following example patterns","documentation and annotations on examples","example usage guides"],"categories":["memory-knowledge","code-generation-editing"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"github_mcp-google-labs-code-stitch-skills__cap_11","uri":"capability://memory.knowledge.design.system.resource.documentation.and.guidelines","name":"design system resource documentation and guidelines","description":"Provides structured reference materials, checklists, style guides, and API documentation in the resources/ directory of each skill, enabling agents to access design system guidelines, component specifications, and best practices during code generation. Resources serve as a knowledge base that agents can query to understand design system constraints, component APIs, styling conventions, and accessibility requirements, improving generation accuracy and consistency.","intents":["I want agents to have access to design system guidelines and component specifications during code generation","I need to document component APIs, styling conventions, and accessibility requirements","I want to provide agents with checklists and best practices to ensure generated code meets standards"],"best_for":["teams maintaining design systems and wanting agent-driven code generation","organizations standardizing on design system documentation","skill developers providing comprehensive reference materials"],"limitations":["resources are static; no mechanism to dynamically update guidelines based on design system changes","agents may not effectively utilize resources if they're poorly organized or documented","no built-in search or indexing; agents must know what resources exist to reference them","resources are skill-specific; no shared resource library across skills"],"requires":["resources/ directory with design system documentation","component specifications and APIs","style guides and coding standards","checklists and best practices","accessibility guidelines"],"input_types":["design system documentation","component specifications","style guides and conventions","accessibility requirements"],"output_types":["agent-accessible reference materials","component API documentation","style guide enforcement","accessibility compliance guidance"],"categories":["memory-knowledge","code-generation-editing"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"github_mcp-google-labs-code-stitch-skills__cap_2","uri":"capability://code.generation.editing.design.to.react.component.code.generation.with.prompt.optimization","name":"design-to-react component code generation with prompt optimization","description":"Transforms UI design data from the Stitch MCP Server into production-ready React components by first optimizing design prompts via the enhance-prompt skill, then generating component code via the react-components skill. The pipeline extracts design semantics (layout, styling, interactivity) from design files and synthesizes React/TypeScript code with proper component structure, prop interfaces, and styling integration, guided by optimized prompts that clarify design intent for the code generation model.","intents":["I want to convert Figma/design tool exports into working React components without manual coding","I need to generate component code that matches design specifications with minimal manual refinement","I want to ensure generated components follow React best practices and are production-ready"],"best_for":["design-to-code teams using Figma or other design tools with Stitch MCP integration","product teams accelerating UI development from design mockups","developers building component libraries from design systems"],"limitations":["requires design data in Stitch MCP Server format; no direct Figma/Sketch file parsing","generated components may require manual refinement for complex interactions or custom logic","styling output depends on enhance-prompt quality; poorly specified designs may generate suboptimal CSS","no built-in support for design tokens or design system variables; requires manual mapping"],"requires":["Stitch MCP Server (latest) with design data loaded","AI coding agent with react-components and enhance-prompt skills installed","Node.js 18+ for component execution and testing","React 18+ project setup"],"input_types":["design metadata from Stitch MCP Server (layout, colors, typography, components)","design prompts or descriptions"],"output_types":["React component files (.tsx/.jsx)","TypeScript prop interfaces","CSS/styled-components styling code","component documentation"],"categories":["code-generation-editing","text-generation-language"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"github_mcp-google-labs-code-stitch-skills__cap_3","uri":"capability://code.generation.editing.multi.page.website.generation.from.design.specifications","name":"multi-page website generation from design specifications","description":"Generates complete multi-page websites (HTML, CSS, JavaScript) from design specifications via the stitch-loop skill, which orchestrates iterative design-to-code transformation across multiple pages. The skill manages page-level decomposition, component reuse across pages, styling consistency, and navigation structure, producing a cohesive website codebase with shared component libraries and unified design system application.","intents":["I want to generate a complete website from a design system specification without building each page individually","I need to ensure consistent styling and component usage across multiple pages","I want to create a website prototype quickly from design mockups for user testing"],"best_for":["design teams prototyping multi-page websites from design systems","product teams building marketing sites or documentation portals from designs","agencies delivering website prototypes to clients rapidly"],"limitations":["requires well-structured design specifications with clear page hierarchies; unstructured designs may generate inconsistent output","no built-in support for dynamic content or backend integration; generated sites are static","navigation and routing are generated as static links; no framework-specific routing (Next.js, React Router) without manual setup","cross-page styling consistency depends on design system quality in source specifications"],"requires":["Stitch MCP Server with multi-page design specifications","stitch-loop skill installed in AI coding agent","design system or component library defined in source specifications","Node.js 18+ for build and preview"],"input_types":["multi-page design specifications from Stitch MCP Server","design system definitions (colors, typography, components)","page-level layout and content specifications"],"output_types":["HTML files for each page","shared CSS stylesheets","component library code","navigation structure and routing configuration","website build artifacts"],"categories":["code-generation-editing","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"github_mcp-google-labs-code-stitch-skills__cap_4","uri":"capability://code.generation.editing.shadcn.ui.component.library.integration.and.guidance","name":"shadcn/ui component library integration and guidance","description":"Provides structured guidance for integrating shadcn/ui components into generated code via the shadcn-ui skill, which includes a component catalog, customization patterns, migration guides, and best practices. The skill enables agents to select appropriate shadcn/ui components for design specifications, apply customization patterns (theming, variant composition), and generate code that leverages the shadcn/ui library instead of building components from scratch, reducing code generation complexity and improving consistency with a widely-used component library.","intents":["I want to use shadcn/ui components in my generated code instead of building custom components","I need guidance on which shadcn/ui component to use for a given design specification","I want to customize shadcn/ui components to match my design system without forking the library"],"best_for":["teams using shadcn/ui as their component library and wanting design-to-code integration","developers building design systems on top of shadcn/ui","product teams accelerating development by leveraging pre-built, accessible components"],"limitations":["requires shadcn/ui to be installed in the target project; no fallback to other component libraries","customization guidance is prescriptive; complex design requirements may not map to shadcn/ui patterns","component catalog is static and requires manual updates when shadcn/ui releases new components","no support for component composition beyond shadcn/ui's built-in composition patterns"],"requires":["shadcn/ui library installed in the React project","shadcn-ui skill installed in AI coding agent","React 18+ and TypeScript","Tailwind CSS (shadcn/ui dependency)"],"input_types":["design specifications from Stitch MCP Server","shadcn/ui component catalog and customization patterns","design system tokens (colors, spacing, typography)"],"output_types":["React component code using shadcn/ui components","customization configuration (CSS, Tailwind classes)","component usage examples","migration guides for existing code"],"categories":["code-generation-editing","memory-knowledge"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"github_mcp-google-labs-code-stitch-skills__cap_5","uri":"capability://text.generation.language.design.system.documentation.generation.from.specifications","name":"design system documentation generation from specifications","description":"Generates comprehensive design system documentation (design-md skill) from design specifications in the Stitch MCP Server, producing markdown files that document design tokens, component definitions, usage patterns, and accessibility guidelines. The skill extracts semantic design information (colors, typography, spacing, components) from design metadata and synthesizes human-readable documentation that serves as a reference for developers and designers, enabling design-to-documentation transformation alongside design-to-code.","intents":["I want to generate design system documentation from my design specifications without manual writing","I need to keep design documentation in sync with actual design specifications","I want to document design tokens, components, and usage patterns for my team"],"best_for":["design teams maintaining design systems and needing documentation automation","product teams documenting design decisions and component usage","organizations standardizing design system documentation across projects"],"limitations":["documentation quality depends on design specification completeness; sparse specifications generate sparse documentation","no support for custom documentation sections or narrative content; output is structured and data-driven","generated markdown may require manual editing for tone, examples, and contextual guidance","no built-in support for versioning or change tracking in documentation"],"requires":["Stitch MCP Server with design specifications loaded","design-md skill installed in AI coding agent","well-structured design metadata (tokens, components, patterns)"],"input_types":["design specifications from Stitch MCP Server","design tokens (colors, typography, spacing, shadows)","component definitions and variants","design patterns and usage guidelines"],"output_types":["markdown documentation files","design token reference tables","component usage examples","accessibility guidelines","design pattern documentation"],"categories":["text-generation-language","code-generation-editing"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"github_mcp-google-labs-code-stitch-skills__cap_6","uri":"capability://image.visual.video.walkthrough.generation.for.component.usage.and.design.patterns","name":"video walkthrough generation for component usage and design patterns","description":"Generates video walkthroughs of components and design patterns via the remotion skill, which synthesizes video content that demonstrates component usage, design system patterns, and interaction flows. The skill uses Remotion (a React-based video generation framework) to programmatically create videos from design specifications and component code, producing shareable video documentation that complements static documentation and code examples.","intents":["I want to create video documentation of components and design patterns without manual video production","I need to demonstrate component interactions and design system usage to stakeholders","I want to generate video walkthroughs of design changes for team communication"],"best_for":["design teams creating component documentation and design system guides","product teams communicating design changes and new features to stakeholders","organizations building comprehensive design system documentation with multimedia content"],"limitations":["video generation is computationally expensive and may require significant processing time","no support for complex interactions or animations beyond what Remotion can synthesize from React components","video output quality depends on component code quality and design specifications","requires Remotion setup and configuration; adds complexity to skill execution environment"],"requires":["Remotion framework installed and configured","remotion skill installed in AI coding agent","React components or design specifications to generate videos from","sufficient computational resources for video rendering"],"input_types":["React component code","design specifications from Stitch MCP Server","interaction flows and animation definitions","design system tokens and styling"],"output_types":["video files (MP4, WebM)","video metadata and transcripts","shareable video links or embeds"],"categories":["image-visual","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"github_mcp-google-labs-code-stitch-skills__cap_7","uri":"capability://text.generation.language.prompt.enhancement.for.improved.code.generation.quality","name":"prompt enhancement for improved code generation quality","description":"Optimizes design prompts and specifications via the enhance-prompt skill to improve downstream code generation quality. The skill analyzes design descriptions, clarifies ambiguous specifications, adds missing context, and structures prompts to maximize code generation model comprehension. This preprocessing step transforms vague or incomplete design specifications into precise, well-structured prompts that guide code generation models toward higher-quality outputs, reducing the need for manual refinement.","intents":["I want to improve the quality of generated code by optimizing design prompts before code generation","I need to clarify ambiguous design specifications so code generation models understand intent","I want to add missing context to design descriptions to guide better code generation"],"best_for":["teams using design-to-code pipelines and wanting to improve output quality","developers working with vague or incomplete design specifications","organizations standardizing on prompt engineering practices for code generation"],"limitations":["prompt enhancement quality depends on the LLM used; different models may produce different optimizations","no guarantee that enhanced prompts will produce better code; depends on downstream code generation model","enhancement process adds latency to the design-to-code pipeline","no feedback loop to measure prompt enhancement effectiveness or iterate on optimization strategies"],"requires":["enhance-prompt skill installed in AI coding agent","design specifications or prompts to enhance","access to LLM for prompt optimization"],"input_types":["design descriptions or prompts","design specifications","design metadata from Stitch MCP Server"],"output_types":["optimized prompts","clarified design specifications","structured design context","prompt quality metrics"],"categories":["text-generation-language","planning-reasoning"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"github_mcp-google-labs-code-stitch-skills__cap_8","uri":"capability://tool.use.integration.external.system.integration.and.workflow.orchestration","name":"external system integration and workflow orchestration","description":"Enables skills to integrate with external systems (APIs, databases, design tools) and orchestrate complex workflows via the standardized scripts/ directory in each skill. Skills can define executable programs (bash, Node.js, Python) that perform network operations, API calls, data transformations, and system integrations, allowing skills to interact with external tools like Figma, GitHub, deployment platforms, and custom backends. The skill framework provides a standard interface for agents to invoke these integration scripts without knowledge of implementation details.","intents":["I want to integrate design-to-code generation with external design tools (Figma, Sketch) or version control systems","I need to orchestrate complex workflows that span multiple systems (design tool → code generation → deployment)","I want to fetch design data from external APIs or databases to feed into code generation"],"best_for":["teams integrating design-to-code pipelines with existing tools and workflows","organizations building end-to-end automation from design through deployment","developers extending skills with custom integrations"],"limitations":["integration scripts are skill-specific; no shared integration framework across skills","external system failures are not gracefully handled; no built-in retry logic or error recovery","no standardized authentication mechanism; each skill must implement its own credential management","scripts/ directory is agent-agnostic but execution environment varies by agent; portability not guaranteed"],"requires":["executable programs in scripts/ directory (bash, Node.js, Python, etc.)","API credentials or authentication tokens for external systems","network connectivity to external systems","agent support for executing arbitrary scripts (security consideration)"],"input_types":["API credentials and configuration","external system data (design files, code repositories, deployment targets)","workflow parameters and orchestration instructions"],"output_types":["integration results (fetched data, API responses)","workflow execution logs","deployed artifacts or updated external systems"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"github_mcp-google-labs-code-stitch-skills__cap_9","uri":"capability://safety.moderation.quality.validation.and.automated.output.checking","name":"quality validation and automated output checking","description":"Enforces output quality through executable validation scripts in the scripts/ directory of each skill, enabling agents to automatically verify generated code, documentation, and other artifacts against success criteria defined in SKILL.md. Validation scripts perform syntax checking, semantic validation, style enforcement, and correctness verification, providing agents with automated feedback on output quality and enabling iterative refinement without manual review.","intents":["I want agents to automatically validate generated code for syntax errors and best practices","I need to enforce quality standards on generated artifacts without manual code review","I want to provide agents with automated feedback so they can iteratively improve outputs"],"best_for":["teams using AI agents for code generation and needing quality assurance","organizations establishing automated quality gates for generated artifacts","developers building skills with strict output requirements"],"limitations":["validation scripts are skill-specific; no shared validation framework across skills","validation can only check syntactic and structural correctness, not semantic correctness or design fidelity","validation scripts must be maintained alongside skill logic; no automatic validation generation","agents may not have permission to execute arbitrary validation scripts (security consideration)"],"requires":["validation scripts in scripts/ directory","success criteria defined in SKILL.md","agent support for executing validation scripts","appropriate linters, type checkers, and validators for the artifact type"],"input_types":["generated artifacts (code, documentation, media)","success criteria and validation rules","validation script definitions"],"output_types":["validation results (pass/fail)","error reports and suggestions for improvement","quality metrics and compliance status"],"categories":["safety-moderation","code-generation-editing"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":49,"verified":false,"data_access_risk":"high","permissions":["Node.js 18+","npm 7+","at least one supported AI coding agent installed (Antigravity, Gemini CLI, Claude Code, or Cursor)","Stitch MCP Server (latest) for most skills to function","adherence to Agent Skills open standard directory structure","SKILL.md file with task description, step-by-step procedures, and success criteria","scripts/ directory with executable validation/integration programs","resources/ directory with reference materials","examples/ directory with syntactically valid reference implementations","syntactically valid reference implementations in examples/ directory"],"failure_modes":["requires agents to be installed and discoverable via standard filesystem paths","no support for agents with custom installation directories or containerized deployments","agent detection is filesystem-based, not API-based, so may fail with non-standard setups","agents must implement parsing logic for SKILL.md format; no guarantee of consistent interpretation across agent implementations","validation scripts are agent-agnostic but may require agent-specific wrappers for execution","no built-in versioning or backward compatibility mechanism for skill format evolution","examples must be manually created and maintained; no automatic example generation","agents may not effectively learn from examples if they're too complex or too simple","examples are static; no mechanism to update examples when best practices evolve","no guarantee that agents will follow example patterns; depends on agent implementation","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.5982526574999905,"quality":0.49,"ecosystem":0.49999999999999994,"match_graph":0.25,"freshness":0.75,"weights":{"adoption":0.25,"quality":0.25,"ecosystem":0.15,"match_graph":0.23,"freshness":0.12}},"observed_outcomes":{"matches":0,"success_rate":0,"avg_confidence":0,"top_intents":[],"last_matched_at":null},"maintenance":{"status":"active","updated_at":"2026-05-24T12:16:22.065Z","last_scraped_at":"2026-05-03T14:23:31.492Z","last_commit":"2026-03-27T17:16:19Z"},"community":{"stars":5150,"forks":618,"weekly_downloads":null,"model_downloads":null,"model_likes":null}},"distribution":{"claim_url":"https://unfragile.ai/submit?claim=mcp-google-labs-code-stitch-skills","compare_url":"https://unfragile.ai/compare?artifact=mcp-google-labs-code-stitch-skills"}},"signature":"Yc30lrC/hC898EquFej6Se20EzsOj/eLhcTQA2ZgZMZo9C0z09g+zbHh8IOpx6rQcHvPlDbs5KWV9ISr64g5Ag==","signedAt":"2026-06-22T09:43:05.366Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/mcp-google-labs-code-stitch-skills","artifact":"https://unfragile.ai/mcp-google-labs-code-stitch-skills","verify":"https://unfragile.ai/api/v1/verify?slug=mcp-google-labs-code-stitch-skills","publicKey":"https://unfragile.ai/api/v1/trust-passport-public-key","spec":"https://unfragile.ai/trust","schema":"https://unfragile.ai/schema.json","docs":"https://unfragile.ai/docs"}}