composable-dxp-claude-marketplace

Brainstorm distinct video concepts for a brief — cinematic videos, brand films, motion accents, and background loops for product websites. Produces 3–5 genuinely different creative directions (not five flavors of one safe idea), each with a premise, the feeling it creates, why it fits the product and placement, and how hard it is to pull off in Higgsfield. Use this skill any time the work starts with "what could this video be" — concepting a hero background loop, a launch film, a section divider, an explainer, or a social spot — even when the user just says "I need a video for X." First stage of the Video Factory chain; hands the chosen concept to video-factory-storyboard. Pulls toward concrete subjects over the drone-over-city / glowing-AI clichés that AI video defaults to.

The main orchestrator for the Slalom Video Factory. Walks a user from "I need a video" to a ready-to-generate prompt package (and, on request, generated clips) via the Frame → Concept → Storyboard → Direct → Generate → Ship pipeline. Diagnoses intent, sequences the craft skills, and pauses for review at each handoff. Use this skill whenever someone wants a cinematic video, a brand film, a motion accent, or a background loop for a product website and hasn't already scoped it down to a single stage — "make me a hero video for X," "I need a background loop," "let's storyboard a launch film." For a single stage, route straight to that craft skill instead.

The motion taste layer for a video — lock one coherent look across every shot before any generation happens. Owns register, pacing, camera language, color and grade, light, grain/texture, and brand fit, so a multi-shot piece reads as one director shot it rather than a pile of clips. The motion analog of design-imagery. Use this skill any time a video needs art direction: setting the visual system for a brand film, keeping a background loop on-brand, grading decisions, or course- correcting clips that came back looking like different videos. Third stage of the Video Factory chain; reads a storyboard, writes a direction spec that video-factory-higgsfield-prompt turns into model parameters and prompt language. Actively rejects the drone-city / glowing-AI motion clichés.

Turn a directed storyboard into a ready-to-run Higgsfield prompt package and (when asked) execute the generations via the Higgsfield MCP. Owns model routing (seedance 2.0 / kling 3.0 default for video; nano banana pro / chatgpt image 2 for keyframes), the image-to-video prompt structure (motion, camera, atmosphere, sound over a start frame), parameter selection (resolution, duration, aspect, sound on/off, genre, end-frame for loops), negatives, retries, and web encoding handoff. Use this skill any time the task is "write the Higgsfield prompt," "generate this in Higgsfield," "what model and settings," "make the background loop," or turning shots into clips. Final craft stage of the Video Factory chain; reads concept + storyboard + direction and produces 40_Prompts.md plus generated keyframes and clips. Keyframe-first, loop- seam-aware, credit-disciplined.

Turn a chosen video concept into a shot-by-shot storyboard ready to generate in Higgsfield — shot list with timing, camera move, framing, transition, and a keyframe-still prompt for every shot (because Higgsfield video models are image-to-video and animate a start frame). Use this skill any time a video idea needs to become an executable shot plan: storyboarding a hero background loop, a brand film, an explainer, a section divider, or a social spot — even when the user says "break this into shots," "plan the shots," or "what's the shot list." Second stage of the Video Factory chain; reads a concept (or a one-line idea) and hands a storyboard to video-factory-creative-direction and video-factory-higgsfield-prompt. Designs loop seams for background video and protects headline safe-areas.

Senior naming expertise. Brainstorms brand names in three categories — literal, abstract, alliterative — with 15+ per category by default. Validates each candidate against existing brand collisions and industry-specific negative connotations (e.g., "Forge" for a law firm reads as "forged documents"). Screens for trademark and domain availability. Use when brainstorming, evaluating, or recommending names for products, companies, features, or offerings.

Assembles the Slalom-branded presenter-mode HTML player into slides/player/. The player is the shippable artifact of a deck. Full presenter mode: current slide + next slide + speaker notes + timer + thumbnail nav + drawing overlay + animation step controls + keyboard shortcuts. Minimalistic chrome, Slalom-branded, single self-contained directory. Reads the slide HTML files and the animations manifest produced by upstream skills.

Stage 3 orchestrator for the Presentation Factory — Compose. Calls presentation-factory-higgsfield and presentation-factory-napkin to generate all images / infographics / icons, then presentation-factory-slide-composer to inline brand tokens into each template and substitute data-slots with content. Output: WIP/[deck]/slides/NN-*.html + deck.html.

The main orchestrator for the Slalom Presentation Factory. Walks the user through the Frame → Draft → Compose → Ship pipeline end-to-end. Uses AskUserQuestion for discrete decisions; conversational input for free-text. Routes through stage orchestrators and pulls specialty skills from the presentation-factory plugin (slalom-brand, higgsfield, napkin, design-layout, animations, html-player, humanize) plus consulting, design, marketing, and brand plugins. Auto-activates when a user runs /presentation-create or starts a new chat in the Presentation Factory Cowork project. Output is the HTML player (canonical); PDF and PPTX are optional derivatives.

Stage 1 orchestrator for the Presentation Factory — Frame. Walks the intake using AskUserQuestion for discrete fields (style, brand, intent, audience, runtime) and conversational input for free-text (purpose, context, source materials). Calls the presentation-factory-slalom-brand for brand ingestion and the presentation-factory-scaffolder agent to set up WIP/[deck]/. Outputs a filled 00_Intake_Brief.md.

Animation choreography for the deck. Owns slide transitions (between slides), build animations (step-by-step reveal within a slide), scroll-triggered reveals (for scroll-export variants), and hover micro-interactions on player chrome. Generates per-slide animation manifests the html-player reads. Reduced-motion aware, GPU-only properties, speaker-controllable.

The Draft stage's authoring skill. Takes the intake brief + research + any source materials and produces 10_Content.md — slide-by-slide content with template assignment, eyebrow/title/body slots, asset prompts (image / infographic / icon), and the italic-Lora emphasis applied where appropriate. Sequences calls to consulting + design + humanize + brand-voice skills.

Layout decisions for a deck. Picks templates from the 25-template catalog per slide, sets visual hierarchy, enforces density discipline (informational vs presentational), and runs the four checks (swap, squint, signature, token) before slide composition. Called by Draft after content is authored, before Compose renders HTML. The skill that turns "here's what the slide says" into "here's how the slide looks."

Stage 2 orchestrator for the Presentation Factory — Draft. Calls presentation-factory-content-author to write the slide-by-slide content; presentation-factory-speaker-notes for notes; consulting + marketing + brand + humanize + design-taste skills for content quality. Pauses for user review at the end. Output: 10_Content.md.

Higgsfield MCP routing for all generative imagery and video in a deck. Routes by asset type: nano banana 2 for stills (default), seedance 2 for video clips, chatgpt image 2 for infographics and any scene that needs text rendered accurately in-image. Owns prompt patterns, dimensions, anti-clichés, retries, and asset naming. Called by Compose stage after content + design-layout decide what each slide needs.

Renders each composed slide HTML to a single PDF page at 1920×1080, then merges them into deck.pdf. Uses headless Chromium via Bash. Validates the output has the right page count and resolution.

Presentation-Factory-specific humanize wrapper over consulting:consulting-humanize. Strips AI tells and consultant jargon, plus adds slide-specific patterns: title brevity, stat-callout register, eyebrow rules, speaker-note voice, and body-prose density limits. Mandatory final pass before any deck is composed, and again after Compose before the deck ships. Operates on 10_Content.md (Draft stage) and on rendered slide HTML (post-Compose).

Napkin.ai routing for data visualizations, conceptual diagrams, frameworks, process flows, value chains, 2x2 matrices, and other structured visuals. Used when the visual is a diagram rather than a photograph and when the visual doesn't need photographic realism. Pairs with presentation-factory-higgsfield (for image-based visuals) and inline SVG (for live data charts).

Converts the deck (HTML slides + PDF) into an editable PowerPoint file where text is editable text frames and assets are movable images. Uses the PowerPoint MCP to create each slide, re-parses the source HTML to extract text + bounding boxes + asset references, and adds them as PPTX elements at the right coordinates. Outputs deck.pptx.

Stage 4 orchestrator for the Presentation Factory — Ship. Calls presentation-factory-html-to-pdf to merge slides into deck.pdf, then presentation-factory-pdf-to-pptx to produce an editable deck.pptx. Calls presentation-factory-promoter agent to move the deck to its canonical home. Archives WIP.

The Slalom brand authority for the Presentation Factory. Owns Slalom-specific tokens, typography (Lora italic titles, Avenir Next subs, Slalom Sans body), the signature italic-emphasis pattern, gradient discipline (strip-only), icon-family rules, and the anti-patterns Slalom decks must avoid. Loaded by Frame for any deck on the Slalom brand (the default). Consumed by content-author, design-layout, and slide-composer as the source of truth for visual register.

The Compose stage's core skill. Reads 10_Content.md and produces one HTML file per slide in WIP/[deck]/slides/, plus a deck.html index that loads every slide for review. For each slide: loads the right template, inlines the brand tokens.css, substitutes every data-slot="..." element with the content from the markdown, and writes the final HTML.

Generates or refines speaker notes for every slide in 10_Content.md. Each slide gets 3-5 sentences the presenter can use as a teleprompter — not the body copy read aloud, but the framing the body copy doesn't include. Runs after content drafting and before composition.

Reads asset prompts from 10_Content.md, routes each to the right generator (Higgsfield for images/video, Napkin.ai for infographics, Flaticon for icons), saves outputs to WIP/[deck]/assets/, and updates 10_Content.md with the resolved local paths. Polls long-running jobs and reports back.

Decides which brand a deck uses (Slalom default, Composable DXP secondary, or Client custom) and prepares the brand assets. For Slalom/Composable, copies the pre-built tokens.css and logo from resources/. For Client, ingests from an attached styleguide PDF or a `.com` crawl, derives tokens, saves to WIP/[deck]/brand/. Outputs: tokens.css + logo.svg in the WIP brand folder.

Health check on an existing Algolia app — settings drift, no-result queries, missing Insights, record-size health, key hygiene, plan-tier fit. Produces a prioritized fix plan. Read-only orchestrator that sequences across the algolia-* skills. Use ~30 days post-bootstrap, then quarterly; also run on engagement handover or before any major relevance change. Doesn't apply changes — produces the plan, then routes per-item.

Stand up Algolia for a new engagement — applications per environment, indices, keys with least-privilege ACLs, settings-as-code repo layout, Insights wiring, and the Contentful indexer scaffold. Orchestrator that sequences across the algolia-* skills end-to-end. Use this skill on greenfield Composable DXP engagements when Algolia is being introduced; do NOT run on an existing Algolia integration with production records (that's a restructure, not a bootstrap — use algolia-audit + algolia-relevance-review instead).

Guided API-key rotation runbook — inventory, create new, deploy, verify, retire old. For staff changes, suspected leaks, scheduled quarterly rotation, or post-incident. Orchestrator that delegates to `algolia-api-keys-security` and enforces the zero-downtime create-deploy-verify-retire sequence per environment. Use this skill when a rotation trigger fires (offboarding, leak, schedule); do NOT use overwrite-in-place patterns.

Guided atomic reindex from the source-of-truth into an Algolia index. Safety rails — non-prod first, settings snapshot, replaceAllObjects, post-reindex validation, run log. Orchestrator that sequences across `algolia-indexing-pipeline`, `algolia-index-design`, `algolia-contentful-integration`, and `algolia-relevance-tuning`. Use this skill for schema migrations, drift repair, or migrations from another search engine; NOT for routine incremental sync (that's handled by webhooks).

Quarterly relevance review using Search Analytics — top searches, no-result queries, low-CTR queries, click position. Proposes synonyms, rules, settings tweaks; defines the A/B plan. Read-only orchestrator that sequences `algolia-analytics-events`, `algolia-relevance-tuning`, and `algolia-personalization-ai`. Use this skill quarterly per index, or on-demand when stakeholders report "search is bad." Produces a proposal package for human approval; doesn't apply changes.

Extract a brand's color, typography, spacing, and elevation system into design-system tokens — semantic naming, HSL channel format, ready for Tailwind/shadcn or Figma variables. Orchestrator that consumes `brand-identity-system` output and feeds `software-engineering-tailwind-tokens` and `software-engineering-shadcn-component`. Use this skill when an identity is locked and ready to flow into code/design-tool variables.

Bootstrap a new Bynder deployment — portal architecture, metaproperty schema, derivatives, permissions, brand guidelines, integrations, with optional Contentful pairing. Orchestrator that sequences across the bynder-* skills end-to-end. Use this skill on greenfield Bynder stand-ups; do NOT run on deployments that already have thousands of assets and active conventions (use bynder-optimization-audit instead).

Audit Bynder Marketplace connector coverage — what's installed, what's actively used, what's stale, and which Tier 1 adoption-driver connectors are missing. Faster, narrower than bynder-optimization-audit — use when the question is specifically about connector fit, not full implementation health. Produces a recommendation list with effort and impact per item.

Run a quarterly Bynder hygiene sweep — audit, then sequentially fix tag/derivative drift, permission decay, broken webhooks, stale brand-guideline pages, and connector misconfigurations. Orchestrator for Slalom-built deployments on a quarterly review cadence (catch drift before it compounds) — distinct from bynder-optimization-audit which is first-time diagnostic on unfamiliar implementations.

Run the four checks (swap, squint, signature, token) plus named-bans audit on a design — produce a tabular Before/After/Why critique. Orchestrator that pulls from `design-taste`, `ai-tells-forbidden-patterns`, `motion-and-transitions`, and `pre-delivery-checklist`. Use this skill on any in-progress design artifact (screen, deck, document, social asset, component) before showing it to the user. Output is named-by-rule, not taste-based.

Design a facilitated session end-to-end — walk the 5 Ps, classify by length/stakes/format, pick the arc, sequence Scan/Focus/Act blocks with named games or modules, pre-design decision moments, anticipate group-dynamics friction, populate the session-design template. Orchestrator that routes to facilitation-meeting-architect, facilitation-workshop-designer, gamestorming-coach, decision-architect, and group-dynamics-coach. Use this skill any time a facilitated session needs structured design — kickoffs, vision workshops, decision sessions, retros, co-creation, planning.

Generate a within-24h workshop recap from session notes, canvas snapshots, or live capture — sponsor-signed, decisions in body, action items with owners and dates, ready to send. Pattern 261 (Make the solution last past Act Day) lives here. Use this skill after every facilitated session that produced decisions or action items; the recap is the artifact that converts a great session into actual change.

Scope a session with the sponsor — Job 1 of Klein & Newman's Six Jobs. Walks the user through outcome definition, out-of-scope naming, participant selection, political-dynamics mapping, and budget questions. Output is a sponsor-aligned scoping brief that names outcomes, decisions already made, what's out of scope, and who must be present. Use this skill BEFORE designing any non-trivial session — scoping is the highest-leverage time spent, and most workshop pain originates from skipping it.

Run a full Innovation Operating System audit across all 7 layers — strategy, governance, discovery, validation, value capture, operating environment, outcomes — with prioritized fix plan. Read-only orchestrator that sequences across every innovation-* skill. Use quarterly or annually, when an innovation effort feels stuck, or before a major restructure. Produces a prioritized remediation plan; doesn't fix issues directly.

Prep for or log a 1:1 with a direct report — opens the running log, surfaces open commitments and standing topics, suggests questions if the slot needs more substance. Orchestrator that combines leadership-people-leader and leadership-meetings-and-cadences with the Manager Tools O3 format. Use this skill before every 1:1 (prep mode) or right after (log mode); follows the running-log-per-direct-report pattern.