SFCC Development MCP Server

import React from 'react'; import { NavLink } from 'react-router-dom'; import SEO from '../components/SEO'; import BreadcrumbSchema from '../components/BreadcrumbSchema'; import StructuredData from '../components/StructuredData'; import CodeBlock, { InlineCode } from '../components/CodeBlock'; import { H1, PageSubtitle } from '../components/Typography'; import { SITE_DATES } from '../constants'; const ModeBadge: React.FC<{ children: React.ReactNode; variant?: 'docs' | 'full' | 'mixed' }> = ({ children, variant = 'docs' }) => { const styles: Record<string, string> = { docs: 'bg-green-100 text-green-700 border-green-300', full: 'bg-blue-100 text-blue-700 border-blue-300', mixed: 'bg-purple-100 text-purple-700 border-purple-300' }; return <span className={`inline-flex items-center gap-1 text-xs font-medium px-2 py-1 rounded-md border ${styles[variant]}`}>{children}</span>; }; const SectionCard: React.FC<{ children: React.ReactNode; gradient?: string; id?: string; title?: string; icon?: string; subtitle?: string }> = ({ children, gradient = 'from-slate-50 via-white to-slate-50', id, title, icon, subtitle }) => ( <section id={id} className="relative mb-14 scroll-mt-20"> <div className={`absolute inset-0 bg-gradient-to-br ${gradient} rounded-2xl`} /> <div className="relative rounded-2xl border border-slate-200/70 bg-white/70 backdrop-blur-sm p-6 md:p-8 shadow-sm"> {title && ( <div className="mb-5"> <h3 className="text-2xl md:text-3xl font-bold tracking-tight text-slate-900 flex items-start gap-3"> <span className="text-xl md:text-2xl" aria-hidden="true">{icon}</span> <span>{title}</span> </h3> {subtitle && <p className="mt-2 text-slate-600 text-sm md:text-base leading-relaxed max-w-3xl">{subtitle}</p>} </div> )} {children} </div> </section> ); const PromptBlock: React.FC<{ prompt: string; intent?: string; }> = ({ prompt, intent }) => ( <div className="rounded-lg border border-slate-200 bg-slate-50 p-4 mb-5"> <p className="text-slate-700 text-sm leading-relaxed"> <span className="font-semibold text-slate-900">Prompt:</span> "{prompt}"{intent && <span className="block mt-1 text-xs text-slate-500">Intent: {intent}</span>} </p> </div> ); const StepsList: React.FC<{ steps: Array<{ label: string; tool?: string; mode?: 'docs' | 'full'; note?: string; }> }> = ({ steps }) => ( <ol className="list-decimal pl-6 space-y-2 text-sm md:text-base"> {steps.map((s, i) => ( <li key={i} className="leading-relaxed"> <span className="font-medium text-slate-800">{s.label}</span> {s.tool && ( <span className="block mt-1 font-mono text-xs md:text-[13px] bg-slate-900 text-slate-100 rounded px-2 py-1 w-fit">{s.tool}</span> )} {s.note && <div className="mt-1 text-xs text-slate-500">{s.note}</div>} {s.mode && <div className="mt-1"><ModeBadge variant={s.mode === 'docs' ? 'docs' : 'full'}>{s.mode === 'docs' ? 'Docs Mode' : 'Full Mode'}</ModeBadge></div>} </li> ))} </ol> ); const ExamplesPage: React.FC = () => { const examplesStructuredData = { "@context": "https://schema.org", "@type": "TechArticle", "headline": "Examples & Workflows - SFCC Development MCP Server", "description": "Prompt-first, end-to-end examples showing how to leverage the SFCC Development MCP Server with AI assistants for real development tasks.", "author": { "@type": "Person", "name": "Thomas Theunen" }, "publisher": { "@type": "Person", "name": "Thomas Theunen" }, "datePublished": SITE_DATES.PUBLISHED, "dateModified": SITE_DATES.MODIFIED, "url": "https://sfcc-mcp-dev.rhino-inquisitor.com/examples/", "educationalLevel": "intermediate", "learningResourceType": "tutorial" }; return ( <div className="max-w-6xl mx-auto px-6 py-8"> <SEO title="Examples & Workflows" description="Prompt-first, end-to-end examples showing how to leverage the SFCC Development MCP Server with AI assistants for real development tasks. Focus on actionable, production-grade outputs." keywords="SFCC examples, Commerce Cloud prompts, MCP workflows, SFCC AI prompts, cartridge generation, log analysis examples" canonical="/examples/" ogType="article" /> <BreadcrumbSchema items={[ { name: "Home", url: "/" }, { name: "Examples", url: "/examples/" } ]} /> <StructuredData data={examplesStructuredData} /> <div className="text-center mb-14"> <H1 id="examples">💡 Prompt-First Examples</H1> <PageSubtitle> Copy a prompt. Paste it into your AI assistant. Get a production-quality answer. Each section shows the exact request, the tool usage behind the scenes, and the kind of response you should expect. </PageSubtitle> <p className="text-sm text-slate-500 max-w-2xl mx-auto">No repetition of feature marketing here—only concrete, minimal, high-signal examples.</p> <p className="mt-4 text-[11px] uppercase tracking-wide text-slate-400">Surface: <strong>36+ specialized tools</strong> (docs, best practices, SFRA, cartridge gen, runtime logs, job logs, system & custom objects, site preferences, code versions)</p> </div> <SectionCard id="explore-class" title="Explore an SFCC Class Quickly" icon="📚" gradient="from-green-50 via-white to-emerald-50" subtitle="Use documentation tools to pull authoritative class context, then build with confidence." > <PromptBlock prompt="Show me everything important about dw.catalog.Product, including methods I actually use in pricing or variation contexts." intent="Documentation deep dive → actionable subset" /> <div className="mb-4 flex flex-wrap gap-2"> <ModeBadge>Docs Mode</ModeBadge> </div> <StepsList steps={[ { label: 'Search for candidate classes (confirmation)', tool: 'search_sfcc_classes {"query": "product"}', mode: 'docs' }, { label: 'Fetch class details', tool: 'get_sfcc_class_info {"className": "dw.catalog.Product"}', mode: 'docs' }, { label: 'Cross reference pricing model', tool: 'get_sfcc_class_info {"className": "dw.catalog.PriceModel"}', mode: 'docs', note: 'Inspect pricing methods' } ]} /> <div className="mt-6"> <p className="text-sm font-semibold text-slate-700 mb-2">Assistant Consolidation (grounded in tool output):</p> <CodeBlock language="markdown" code={`### Focused Summary: dw.catalog.Product Pulled full class definition (get_sfcc_class_info → dw.catalog.Product) and reduced to what typically drives storefront pricing / availability / variation logic. Identity & Routing: • ID, name, pageURL for canonical + SEO. • onlineFlag vs online (flag is raw, online applies scheduling window). Pricing: • priceModel.getPrice().value / currencyCode for base price display. • Use getPriceModel(optionModel) when option selections adjust pricing. • Guard: priceModel can be null on offline or misconfigured variants. Availability: • availabilityModel.isInStock() and availabilityModel.getATS()?.value for inventory messaging. • Avoid deprecated isAvailable()/getAvailableFlag(); rely on availabilityModel. Variation Handling: • isMaster() / isVariant() to branch logic. • variationModel.getVariants() (iterate then filter by onlineFlag). • Image fallback: getImage('small', 0) (uses variant images, then master fallback); avoid deprecated no-arg getImage(). Navigation / Category Context: • onlineCategories preferred for building live nav trees; allCategories includes offline / non-site assignments. Common Pitfalls & Mitigations: 1. Null priceModel → defensive optional chaining. 2. Using deprecated availability helpers → switch to availabilityModel. 3. Assuming images exist → null check getImage before URL access. 4. Iterating variants without online gating → filter by v.onlineFlag. Minimal Access Snippet: \`var pm = p.getPriceModel();\nvar price = pm ? pm.getPrice().value : null;\nvar inStock = p.getAvailabilityModel().isInStock();\nif (p.isMaster()) {\n var variants = p.getVariationModel().getVariants().toArray().filter(v => v.onlineFlag);\n}\` `} /> </div> </SectionCard> <SectionCard id="focused-class-exploration" title="Focused Class Exploration with Filtering" icon="🔍" gradient="from-emerald-50 via-white to-teal-50" subtitle="Use enhanced filtering and search to find exactly what you need from SFCC classes." > <PromptBlock prompt="I need to understand only the getter methods in dw.catalog.Product that relate to pricing or availability. Don't show me everything." intent="Targeted exploration with precise filtering" /> <div className="mb-4 flex flex-wrap gap-2"> <ModeBadge>Docs Mode</ModeBadge> </div> <StepsList steps={[ { label: 'Show only methods with price-related terms', tool: 'get_sfcc_class_info {"className": "dw.catalog.Product", "includeProperties": false, "includeConstants": false, "includeDescription": false, "includeInheritance": false, "search": "price"}', mode: 'docs' }, { label: 'Show only methods with availability terms', tool: 'get_sfcc_class_info {"className": "dw.catalog.Product", "includeProperties": false, "includeConstants": false, "includeDescription": false, "includeInheritance": false, "search": "availab"}', mode: 'docs' }, { label: 'Compare with full getter methods', tool: 'get_sfcc_class_info {"className": "dw.catalog.Product", "includeProperties": false, "includeConstants": false, "includeDescription": false, "includeInheritance": false, "search": "get"}', mode: 'docs', note: 'See complete getter landscape' } ]} /> <div className="mt-6"> <p className="text-sm font-semibold text-slate-700 mb-2">Focused Results (price-related methods only):</p> <CodeBlock language="markdown" code={`### dw.catalog.Product - Price Methods Only **Methods:** • getPriceModel() : PriceModel - Returns pricing model for current product • getPriceModel(ProductOptionModel) : PriceModel - Returns pricing with option adjustments **Key Benefits:** - Clean, targeted view without noise from 50+ other methods - Focus on exactly what's needed for pricing logic - Search functionality quickly surfaces relevant APIs`} /> <div className="mt-4 p-3 bg-blue-50 rounded-lg"> <p className="text-sm text-blue-800"><strong>Pro Tip:</strong> Use <code>search</code> parameter with partial terms like "get", "set", "price", "availab" to quickly find relevant methods. Combine with <code>includeProperties: false</code> to focus only on methods.</p> </div> </div> </SectionCard> <SectionCard id="controller-from-docs" title="Generate a Controller from Documentation Context" icon="🛠️" gradient="from-blue-50 via-white to-indigo-50" subtitle="Documentation → intent → runnable controller with guard rails." > <PromptBlock prompt="Using SFRA patterns, create a Product-Show like controller action that handles missing products, logs errors, and renders a view model with price + availability." intent="Transform docs into production-safe code" /> <ModeBadge>Docs Mode</ModeBadge> <div className="mt-4 text-sm text-slate-600">Assistant internally referenced previously fetched class info; no additional tool calls required now.</div> <div className="mt-6"> <CodeBlock language="javascript" code={`// app_storefront_base/cartridge/controllers/CustomProduct.js 'use strict'; var server = require('server'); var ProductMgr = require('dw/catalog/ProductMgr'); var Logger = require('dw/system/Logger'); var cache = require('*/cartridge/scripts/middleware/cache'); server.get('Show', cache.applyDefaultCache, function (req, res, next) { var pid = req.querystring.pid; if (!pid) { res.setStatusCode(400); res.render('error/badrequest'); return next(); } try { var product = ProductMgr.getProduct(pid); if (!product || !product.onlineFlag) { res.setStatusCode(404); res.render('error/notfound'); return next(); } var availability = product.getAvailabilityModel(); var priceModel = product.getPriceModel(); var viewModel = { id: product.ID, name: product.name, price: priceModel ? priceModel.getPrice().value : null, currency: priceModel ? priceModel.getPrice().currencyCode : null, inStock: availability.isInStock(), ats: availability.getATS && availability.getATS().value, images: product.getImages('small', 0).toArray().map(function(i){return i.URL;}) }; res.render('product/customProduct', viewModel); } catch (e) { Logger.error('CustomProduct.Show error: {0}', e.message); res.setStatusCode(500); res.render('error/general'); } return next(); }); module.exports = server.exports();`} /> </div> </SectionCard> <SectionCard id="cartridge-generation" title="Scaffold a Feature Cartridge" icon="📦" gradient="from-purple-50 via-white to-fuchsia-50" subtitle="Go from blank folder to structured integration base." > <PromptBlock prompt="Generate a cartridge structure for a shipment tracking integration using hooks + a service layer and note follow-up tasks." intent="Structural generation + next actions" /> <div className="mb-4 flex flex-wrap gap-2"><ModeBadge>Docs Mode</ModeBadge></div> <StepsList steps={[ { label: 'Generate skeleton', tool: 'generate_cartridge_structure {"cartridgeName": "int_tracking"}', mode: 'docs' }, { label: 'Request best practice follow-up', tool: 'get_best_practice_guide {"guideName": "cartridge_creation"}', mode: 'docs', note: 'Assistant merges guidance' } ]} /> <div className="mt-6 space-y-8"> <div> <p className="text-sm font-semibold text-slate-700 mb-2">Generated Tree (excerpt):</p> <CodeBlock language="text" code={`int_tracking/ └── cartridge/ ├── controllers/ │ └── Tracking.js ├── scripts/ │ ├── hooks/ │ │ └── shipment.js │ └── services/ │ └── TrackingService.js ├── templates/default/tracking/ │ └── trackingstatus.isml └── static/default/js/tracking/`} /> </div> <div> <p className="text-sm font-semibold text-slate-700 mb-2">Assistant Adds Next Steps:</p> <CodeBlock language="markdown" code={`1. Configure LocalServiceRegistry service: tracking.api 2. Implement hooks: updateShippingInfo, exportStatus 3. Add ISML partial to PDP for shipment ETA 4. Add logging category: TrackingIntegration 5. Write integration test harness for mocked API responses `} /> </div> </div> </SectionCard> <SectionCard id="full-mode-diff" title="Enhance with Live Logs (Full Mode)" icon="🧪" gradient="from-blue-50 via-white to-sky-50" subtitle="Same style of prompt—but now the assistant executes log tooling to ground its answer." > <PromptBlock prompt="Checkout is intermittently failing with payment declines—identify root causes from recent logs and propose concrete mitigations." intent="Diagnostic + remediation planning" /> <div className="flex flex-wrap gap-2 mb-4"><ModeBadge variant="full">Full Mode</ModeBadge></div> <StepsList steps={[ { label: 'Daily context snapshot', tool: 'summarize_logs {"date": "20250916"}', mode: 'full', note: 'High-level counts first' }, { label: 'Attempt error retrieval', tool: 'get_latest_error {"limit": 15}', mode: 'full', note: 'May return none → fall back to warnings' }, { label: 'Focus on recurring warning pattern', tool: 'search_logs {"pattern": "cookie_hint", "logLevel": "warn", "limit": 5}', mode: 'full', note: 'Isolate noisy asset issue' }, { label: 'Cross-check security guidelines (token lifecycle / logging hygiene)', tool: 'get_best_practice_guide {"guideName": "security"}', mode: 'docs' } ]} /> <div className="mt-6"> <p className="text-sm font-semibold text-slate-700 mb-2">Assistant Narrative (validated with real tool set):</p> <CodeBlock language="markdown" code={`### Live Log Triage (No Errors Detected Today) 1. summarize_logs → Errors: 0, Warnings: 27, no debug/info noise. Healthy baseline but warning cluster present. 2. get_latest_error → Returned none (expected—error file absent). Pivoting to warnings. 3. search_logs (pattern=cookie_hint, level=warn) → Multiple repeats: "Content asset with ID cookie_hint is offline" across distinct request contexts. Assessment: • Repeated offline content asset warnings inflate noise floor; risk of masking future meaningful warnings. • No payment/auth related issues surfaced; current checkout remediation unnecessary. Recommended Actions: • Restore or intentionally remove the cookie_hint asset; if intentionally removed, suppress via conditional include. • Add lightweight monitoring rule: if warning count for a single asset > X per hour, raise housekeeping task instead of polluting logs. • Keep error path watch in place; re-run summarize_logs after remediation to confirm warning reduction. If Errors Had Been Present (Pattern Skeleton): • Step 1 would still be summarize_logs for counts → then targeted search_logs for signatures (e.g., AUTH_TIMEOUT, TOKEN_EXPIRED) followed by best practice cross-reference. `} /> </div> </SectionCard> <SectionCard id="job-log-analysis" title="Investigate a Failing Job" icon="🧵" gradient="from-amber-50 via-white to-yellow-50" subtitle="Targeted job log inspection with focused summarization." > <PromptBlock prompt="Analyze the latest CatalogFeed job execution – summarize step durations, failures, and recommend optimizations." intent="Temporal + failure analysis" /> <div className="flex flex-wrap gap-2 mb-4"><ModeBadge variant="full">Full Mode</ModeBadge></div> <StepsList steps={[ { label: 'Discover job logs (filtered)', tool: 'search_job_logs_by_name {"jobName": "CatalogFeed", "limit": 5 }', mode: 'full', note: 'Find specific job log filenames' }, { label: 'List latest job logs (context)', tool: 'get_latest_job_log_files {"limit": 5 }', mode: 'full', note: 'Recency overview (no name filter)' }, { label: 'Fetch recent entries', tool: 'get_job_log_entries {"jobName": "CatalogFeed", "limit": 50, "level": "all" }', mode: 'full', note: 'Multi-level snapshot' }, { label: 'Search for errors', tool: 'search_job_logs {"pattern": "ERROR", "jobName": "CatalogFeed", "limit": 20 }', mode: 'full', note: 'If none → treat as healthy case' }, { label: 'Execution summary', tool: 'get_job_execution_summary {"jobName": "CatalogFeed"}', mode: 'full' } ]} /> <div className="mt-6"> <CodeBlock language="markdown" code={`### CatalogFeed Execution Analysis (Validated Flow) Flow Used: 1. search_job_logs_by_name → confirm job exists. 2. get_latest_job_log_files → establish relative recency. 3. get_job_log_entries (limit 50, all levels) → capture operational window. 4. search_job_logs (pattern=ERROR) → branch: errors vs healthy. 5. get_job_execution_summary → aggregate timing + status. Illustrative Degraded Scenario (for instructional value): Duration: 7m42s (↑18% vs rolling average) Step Highlights: • LOAD_PRODUCTS: 2m10s (warning — 312 fallback fetches suggests cache miss storm) • EXPORT: 15 socket timeout retries before success (dominant latency source) Primary Bottleneck: External export endpoint + un-jittered linear retry compounding latency. Recommendations (priority order): 1. Reduce batch size 200→120 to shrink payload variance. 2. Exponential backoff + jitter (≤800ms, cap 6) for EXPORT retries. 3. Parallelize transform stage (2 workers) given idle CPU. 4. Emit structured metrics: step_duration_ms, retry_count to confirm remediation. Healthy Case Note: If error search returns 0 (common for maintenance/cleanup jobs) emit concise summary only and skip remediation suggestions. `} /> </div> </SectionCard> <SectionCard id="system-object-attributes" title="Surface High-Value Custom Attributes" icon="🧬" gradient="from-rose-50 via-white to-pink-50" subtitle="Focus on signal: only the attributes you will likely act on in code." > <PromptBlock prompt="List only non-system custom Product attributes that impact pricing, display, or integrations—include access examples." intent="Selective attribute curation" /> <div className="flex flex-wrap gap-2 mb-4"><ModeBadge variant="full">Full Mode</ModeBadge></div> <StepsList steps={[ { label: 'Search attribute definitions', tool: 'search_system_object_attribute_definitions {"objectType": "Product", "searchRequest": {"query": {"match_all_query": {}}}}', mode: 'full' }, { label: 'Filter & rank', note: 'Assistant filters system vs custom, groups by usage relevance' }, ]} /> <div className="mt-6"> <CodeBlock language="markdown" code={`### Curated Product Attribute Focus (Grounded in Live Metadata) Source: search_system_object_attribute_definitions → 113 total attributes (first 30 sampled). We extracted *custom* (system=false) fields + a few pivotal system ones (clearly flagged) that materially influence merchandising, variation, UX, integration, or feed logic. Purely operational / audit (UUID, creationDate) or low-impact marketing fluff excluded. Legend: • Cat: Category of impact (Disp = Display/UX, Var = Variation, Facet = Faceting/Search, Int = Integration/Channel, Enr = Enrichment, Dim = Dimensions/Shipping) • MV = multi-value (enum_of_string or multi_value_type true) • SYS = system attribute retained for high value (otherwise we focus on custom) | Attribute | Cat | Type | Flags | Rationale | Access Snippet | |-----------|-----|------|-------|-----------|----------------| | brand (SYS) | Facet/Disp | string | visible | Drives brand badges & facet grouping; high-frequency filter. | \`p.brand\` (system field, not in p.custom) | | color | Var | string | custom | Variation presentation / swatch resolution; stored custom but effectively a core merchandising dimension. | \`p.custom.color\` | | availableForInStorePickup | Int | boolean | custom, site-agnostic | Gating for store pickup workflows & conditional UI messaging. | \`if (p.custom.availableForInStorePickup) showPickup();\` | | batteryLife | Enr/Disp | string | custom, visible | PDP spec table + comparison view; enriches SEO content. | \`p.custom.batteryLife\` | | batteryType | Enr/Disp | string | custom, visible | Hardware spec grouping; feed enrichment for certain channels. | \`p.custom.batteryType\` | | bootType | Facet | enum_of_string | custom, MV | Multi-select refinement facet; influences search narrowing logic. | \`p.custom.bootType && p.custom.bootType.toArray()\` | | bottomType | Facet | enum_of_string | custom, MV | Apparel categorization & layered navigation. | \`(p.custom.bottomType||[]).toArray()\` | | consoleWarranty | Int/Disp | string | custom | Extended service offering display + export to warranty provider feed. | \`p.custom.consoleWarranty\` | | digitalCameraFeatures | Facet/Enr | enum_of_string | custom, MV, visible | Feature badges + facet refinement; high cardinality set. | \`p.custom.digitalCameraFeatures?.toArray()\` | | digitalCameraPixels | Disp/Enr | string | custom, visible | Marketing resolution highlight; conditional comparison logic. | \`p.custom.digitalCameraPixels\` | | digitalCameraType | Facet | string | custom, visible | Primary camera classification facet. | \`p.custom.digitalCameraType\` | | digitalCameraWarranty | Int | string | custom | Feed & PDP legal/service disclosure. | \`p.custom.digitalCameraWarranty\` | | dimDepth | Dim | string | custom, visible | Shipping dimensional weight calc if external DIM service absent. | \`Number(p.custom.dimDepth)\` | | dimHeight | Dim | string | custom, visible | Parcel sizing / volumetric charge triggers. | \`Number(p.custom.dimHeight)\` | | dimWeight | Dim | string | custom, visible | Fallback for missing system weight or override scenario. | \`Number(p.custom.dimWeight)\` | | dimWidth | Dim | string | custom, visible | Combined with others for cubic volume. | \`Number(p.custom.dimWidth)\` | | displaySize | Disp | string | custom, visible | Prominence in electronics PDP hero section + compare grid. | \`p.custom.displaySize\` | | facebookEnabled (SYS) | Int | boolean | system, site_specific | Channel feed toggle – dictates inclusion in FB catalog export. | \`p.facebookEnabled\` | | gameGenre | Facet | enum_of_string | custom, MV, visible | Multi-select discovery facet for gaming taxonomy. | \`p.custom.gameGenre?.toArray()\` | | gameRating | Facet/Compliance | string | custom, visible | Age rating badge + compliance gating (e.g., age verification). | \`p.custom.gameRating\` | | gameSystemType | Facet | string | custom, visible | Platform segmentation (PS/Xbox/Nintendo) – major navigation axis. | \`p.custom.gameSystemType\` | | gpsFeatures | Facet/Enr | enum_of_string | custom, MV, visible | Feature-level refinement & badge cluster. | \`p.custom.gpsFeatures?.toArray()\` | | Wool | Enr/Disp | string | custom, localizable, visible | Care instruction overlay & localized PDP sustainability note. | \`p.custom.Wool\` | Why exclude others like EAN / UPC (system, typically handled upstream) or UUID (audit only)? They rarely drive conditional storefront logic once imported and are better treated as pass-through feed fields. Implementation Notes: 1. Multi-value handling: Always null-check before toArray(); some may be undefined on non-merchandised variants. 2. Numeric coercion: DIM fields are strings – coerce with Number() and guard NaN before calculations. 3. Site specificity: facebookEnabled is site_specific – avoid caching decisions across sites without keying by site ID. 4. Variation interplay: color selection should reference variationAttribute values; treat p.custom.color as display fallback not authoritative source. 5. Performance: When projecting many attributes into a model, prefer a single transform object rather than repeated p.custom dereferencing inside loops. Minimal Extraction Snippet: \`var c = p.custom;\nvar model = {\n brand: p.brand,\n color: c.color,\n pickupEligible: !!c.availableForInStorePickup,\n dims: ['dimHeight','dimWidth','dimDepth','dimWeight'].reduce(function(acc,k){var v=c[k]; if(v && !isNaN(v)) acc[k.replace('dim','').toLowerCase()]=Number(v); return acc;}, {}),\n features: (c.digitalCameraFeatures && c.digitalCameraFeatures.toArray()) || [],\n care: c.Wool,\n fbActive: p.facebookEnabled\n};\nif (c.bootType) model.bootType = c.bootType.toArray();\nif (c.gameGenre) model.gameGenre = c.gameGenre.toArray();\` `} /> </div> </SectionCard> <SectionCard id="site-preference-search" title="Discover Site Preferences Safely" icon="⚙️" gradient="from-indigo-50 via-white to-purple-50" subtitle="Locate configuration switches before you hardcode assumptions." > <PromptBlock prompt="Find all site preferences in the checkout group related to tax or shipping so I can reference them defensively in code." intent="Configuration discovery" /> <div className="flex flex-wrap gap-2 mb-4"><ModeBadge variant="full">Full Mode</ModeBadge></div> <StepsList steps={[ { label: 'List groups (if unknown)', tool: 'search_system_object_attribute_groups {"objectType": "SitePreferences", "searchRequest": {"query": {"match_all_query": {}}}}', mode: 'full', note: 'Optional first pass – discover group IDs' }, { label: 'Search preferences in group', tool: 'search_site_preferences {"groupId": "checkout", "searchRequest": {"query": {"text_query": {"fields": ["id","display_name","description"], "search_phrase": "tax"}}, "count": 50}}', mode: 'full' }, { label: 'Broaden to shipping', tool: 'search_site_preferences {"groupId": "checkout", "searchRequest": {"query": {"text_query": {"fields": ["id","display_name","description"], "search_phrase": "ship"}}, "count": 50}}', mode: 'full' } ]} /> <div className="mt-6"> <CodeBlock language="markdown" code={`### Interpreting Results Each preference includes: id, display_name, value_type, (optionally) value definition. Security: Password-type preference values are masked by default. Usage Pattern: \`var prefs = Site.getCurrent().getPreferences().getCustom();\nvar enableAltTax = prefs.enableAlternativeTaxEngine;\nif (enableAltTax) { /* branch logic */ }\` Defensive Access Tips: 1. Guard optional boolean flags with !! to normalize (e.g., !!prefs.enableGiftWrap) 2. Never log preference values directly (especially STRING/LONG_TEXT types containing tokens) 3. Document critical preference dependencies in controller/module JSDoc `} /> </div> </SectionCard> <SectionCard id="custom-object-attributes" title="Target Custom Object Attributes" icon="🧱" gradient="from-teal-50 via-white to-cyan-50" subtitle="Search attributes precisely instead of browsing blindly." > <PromptBlock prompt="For custom object type Global_String, list only searchable non-system attributes and show how to query one in script." intent="Metadata → usage" /> <div className="flex flex-wrap gap-2 mb-4"><ModeBadge variant="full">Full Mode</ModeBadge></div> <StepsList steps={[ { label: 'Query searchable attributes', tool: 'search_custom_object_attribute_definitions {"objectType": "Global_String", "searchRequest": {"query": {"bool_query": {"must": [{"term_query": {"fields": ["searchable"], "operator": "is", "values": ["true"]}}]}}, "count": 100}}', mode: 'full' }, { label: '(Optional) broaden if empty', note: 'Fallback: use match_all_query then client-side filter' } ]} /> <div className="mt-6"> <CodeBlock language="markdown" code={`### Usage Extraction Attribute Sample: altValueMapping (type=string, searchable=true) Script Access: \`var CustomObjectMgr = require('dw/object/CustomObjectMgr');\nvar co = CustomObjectMgr.getCustomObject('Global_String', key);\nvar mapped = co && co.custom.altValueMapping;\` Guidelines: - Null-check custom object fetch before dereferencing .custom - Prefer specific term_query filtering serverside rather than large client-side scans - If attribute becomes critical path, add defensive logging (category=MetadataAccess, level=debug) selectively (avoid value leakage) `} /> </div> </SectionCard> <SectionCard id="micro-job-log-triage" title="Micro Job Log Triage" icon="🩺" gradient="from-red-50 via-white to-rose-50" subtitle="Fast 3-step health read of a job without noise." > <PromptBlock prompt="Give me a 3-step minimal health read for the nightly InventorySync job and only call tools you truly need." intent="Minimal observation pipeline" /> <div className="flex flex-wrap gap-2 mb-4"><ModeBadge variant="full">Full Mode</ModeBadge></div> <StepsList steps={[ { label: 'Confirm presence', tool: 'search_job_logs_by_name {"jobName": "InventorySync", "limit": 3 }', mode: 'full' }, { label: 'Get recent entries (tail)', tool: 'get_job_log_entries {"jobName": "InventorySync", "limit": 40 }', mode: 'full', note: 'Mixed levels in single job file' }, { label: 'Execution summary', tool: 'get_job_execution_summary {"jobName": "InventorySync"}', mode: 'full', note: 'Roll-up status + durations' } ]} /> <div className="mt-6"> <CodeBlock language="markdown" code={`### Minimal InventorySync Health Snapshot Outcome: SUCCESS (duration 2m14s, 0 errors, 3 warnings) Warnings Focus: • 2x RETRY: External API 429 backoffs (acceptable within thresholds) • 1x Deprecated attribute reference (non-blocking; schedule fix) Next Actions (Only If Persistent): 1. Track retry ratio vs baseline (add metric instrumentation) 2. Replace deprecated attribute before Q4 freeze `} /> </div> </SectionCard> {/* End-to-End section removed per request */} <SectionCard id="prompt-patterns" title="Prompt Patterns & Anti-Patterns" icon="🧭" gradient="from-slate-50 via-white to-slate-100" subtitle="Tight prompts yield grounded, production-usable outputs." > <div className="grid md:grid-cols-2 gap-8"> <div> <h4 className="font-semibold text-slate-800 mb-2">Effective Patterns</h4> <ul className="space-y-2 text-sm list-disc pl-5"> <li><strong>Role + Scope:</strong> "Act as SFRA dev – generate..."</li> <li><strong>Constraint:</strong> "Only include attributes affecting pricing."</li> <li><strong>Mode Awareness:</strong> "Use live logs to confirm before proposing fixes."</li> <li><strong>Transformation:</strong> "Summarize for junior dev handoff."</li> </ul> </div> <div> <h4 className="font-semibold text-slate-800 mb-2">Anti-Patterns</h4> <ul className="space-y-2 text-sm list-disc pl-5"> <li>"Explain everything about SFCC" (Too broad)</li> <li>"Fix checkout" (No signal / context)</li> <li>Omitting objective (no success definition)</li> <li>Forgetting mode capabilities (asks for logs in docs-only)</li> </ul> </div> </div> <div className="mt-6"> <p className="text-sm font-semibold text-slate-700 mb-2">Prompt Refinement Example:</p> <CodeBlock language="markdown" code={`Weak: "Help with product page" Improved: "Generate an SFRA controller extension to enrich Product-Show with cached ATS and badge if sustainabilityRating >=3. Provide only changed code + template snippet."`} /> </div> </SectionCard> <div className="text-center mt-20"> <p className="text-lg text-slate-700 mb-6 font-medium">Ready to try these yourself?</p> <div className="flex flex-col sm:flex-row gap-4 justify-center"> <NavLink to="/ai-interfaces/" className="group no-underline hover:no-underline bg-gradient-to-r from-blue-600 to-purple-600 text-white px-8 py-4 rounded-xl font-semibold text-lg shadow-lg hover:shadow-xl transition-all duration-300 transform hover:-translate-y-1"> Configure Your AI Client <span className="ml-2 group-hover:translate-x-1 inline-block transition-transform">→</span> </NavLink> <NavLink to="/tools/" className="no-underline hover:no-underline border-2 border-slate-300 text-slate-700 px-8 py-4 rounded-xl font-semibold text-lg hover:border-blue-500 hover:text-blue-600 transition-all duration-300"> Browse All Tools </NavLink> </div> </div> </div> ); }; export default ExamplesPage;