bctechdays-presentation

name: bctechdays-presentation description: Build HTML presentation decks styled like the BC TechDays 2026 brand — a LIGHT cool-grey canvas, RED Montagu Slab slab-serif titles (#F7514F), Corbel body, full-bleed indigo/coral/magenta colour "moment" panels, the real BC TechDays scan-line "glitch wave" + wireframe-roster motif and torn-ticket "BC/NAV TECH DAYS 2026 · mibuso.com" logo (both inlined from the official PowerPoint), and the full pattern engine (claim-stack prov, centered tc-context, six-pillar tc-map, MCP hub, animated bridge, df-day timeline, htrack stage-tracker, agents-team grid, pb-hero playbook bridge, three-node storylines, pipeline analyzer, hub-and-spokes CSP, hands-up audience cards, replay overlay for embedded chat replays, inline presenter notes, View Transitions / element morph, a persistent dogsled trail progress tracker with build-trail overview morph, print-pdf export mode, click-to-zoom vocabulary cards, terminology grid, strap list, photo/instruction splits, image galleries, a retro C/AL window skin, in-flow replay row, extended chip types, SVG zoom, lavender Q&A/Thank-You enders). Trigger whenever the user asks for a BC TechDays deck, BC TechDays slides, a BC TechDays-branded presentation, or a deck for a BC TechDays / mibuso session or talk. Business Central / BC / NAV conference talks default to this look.

BC TechDays presentation skill

Use this skill any time the user wants an HTML presentation that should match the BC TechDays 2026 brand. The design tokens below were sampled directly from the official BCTechDays2026 PowerPoint template (theme colours, embedded fonts, and the logo/motif artwork) — not from a guess and not from the marketing site's raw CSS (which exposes a different lime token that is not how the presentation brand actually reads). Don't substitute values.

This skill shares the proven deck engine of the iFacto presentation skill (keyboard nav, step reveals, typewriter, slide picker with topic separators, present mode) but is a completely different brand.

How to use this skill

1. Open assets/template.html. It's a working multi-slide deck with every pattern below wired up, plus keyboard nav and the wave/mesh motif generator. Copy it, then swap content.
2. The demo content is generic placeholder. Replace it with the real talk; don't rewrite the CSS.
3. Delete patterns you don't need. Keep the rhythm (see "Slide rhythm" below).
4. Save the deck to the user's workspace folder, not the skill folder.

Slide density — minimum text on every slide

The presenter is the verbal layer; the slide is a visual anchor, not a teleprompter.

• Readable in under 5 seconds without the presenter speaking.
• One strong sentence > a paragraph. A 1–3 word label > a caption.
• No bulleted prose. Card labels are 1–3 words ("Orchestration", not a sentence).
• A headline + one accent word + the motif is enough for most slides.
• Over ~15 visible words → it's two slides. Split it.

If a pattern in template.html shows denser demo text, strip it down before shipping.

Brand tokens — non-negotiable

Sampled from the official .pptx. The deck is light-dominant: a cool off-white canvas, punctuated by full-bleed colour panels for the "moment" slides. Titles are RED #F7514F (the .pptx "main title" colour); indigo #6165E6 is the dominant panel / component colour. The rest are a deliberate carnival used for categorical meaning (different cards / tracks / sections), not random decoration.

Fonts (the families the official deck actually embeds):

• Montagu Slab — display headlines (h1–h3), Bold/Black. A slab serif. This is the BC TechDays voice. Loaded from Google Fonts. Enforced via !important so any <h1>/<h2>/<h3> inherits it.
• Corbel — body, captions, lead, the slide canvas default. A humanist sans (installed on Windows/Office machines). The stack is "Corbel","Hanken Grotesk",system-ui,…; Hanken Grotesk (Google) is the portable fallback so the deck still looks right on machines without Corbel.
• Barlow Condensed — eyebrows, pager, picker chrome, kickers, numbers, the logo lockup, .df-tag, .cnum, .secnum. Heavy condensed grotesque, uppercase, tracked — it echoes the "BC TECH DAYS" wordmark.

Do not use lime #CCFF00, dark/near-black backgrounds, or Poppins/Inter/Barlow-as-headline. Those were an earlier wrong read of the brand. The presentation brand is light + slab serif + indigo.

Type scale

h1 (hero/statement)  clamp(46px,5.6vw,88px)  Montagu Slab 700, line-height .98
h2 (slide headline)  clamp(48px,6vw,92px)    Montagu Slab 700, line-height 1.04
h3 (card title)      clamp(28px,2.4vw,40px)  Montagu Slab 600, line-height 1.2
h4                   clamp(22px,1.7vw,28px)  Montagu Slab, line-height 1.3
.eyebrow             clamp(14px,1.05vw,17px) Barlow Condensed 700, UPPERCASE, tracked .13em
.lead                clamp(30px,2.3vw,42px)  Corbel/Hanken 500
p                    clamp(20px,1.55vw,28px) Corbel/Hanken 400
.pager / .topnav     14px / 15px             Barlow Condensed, tracked

Body uses bold inside light text to emphasise 1–3 keywords mid-sentence. Never bold a whole sentence.

Title colour & the optional accent word

• On the light canvas, headlines (h1/h2/h3) are RED #F7514F — that's the .pptx main-title colour. (They are not black-with-a-coloured-word, and not purple/indigo.) Card-title divs stay black.
• On a colour panel (.slide.fill), headlines are white.
• The accent span <span class="g">word</span> makes ONE word stand out: on light it's ink/black (a dark word inside the red title — two-tone); on a panel it flips to yellow. Use it sparingly or not at all — a fully-red title is the default and is fine.

Never colour more than one word/phrase per headline.

Slide patterns in the template

This deck carries the full pattern engine — every pattern below is wired in template.html, with <style> defs in the head and a working demo slide in the body. Copy and adapt the ones you need; don't rewrite the CSS.

Title slides

1. .title-photo-slide — full-bleed real artwork from the official PowerPoint. Inline an exported JPG as <img class="title-photo"> for a true single-file deck, or use .title-img-slide with background-image:url('title-session2.jpg') for an externally-referenced file. The deck ships with one of each. Pasted-in-chat images can't be used directly — save them as files first.
2. Animated hero (.slide without .light/.fill) — a "Welcome" slide with the .hero-bg orb backdrop (glow1 indigo / glow2 yellow / glow3 green, all blurred and floating), the .mesh-plane wireframe, and a .wave.hero corner accent. The hero-bg orbs are visible on this slide; on other slides they get neutralised by a brand override.

Light-canvas content patterns (.slide.light)

3. Provocation / claim stack (.slide.light.prov) — 3–4 headlines that type/reveal on click (.prov-line.step.typed). Claim → reframe → implication → call-to-action. Often paired with an image on the right. Background orbs are dimmed so the text dominates.
4. Centered intro (.slide.light.tc-context) — centered headline + lead + an optional reference link. The section opener / "before we go further" moment.
5. Six-pillar toolchain map (.slide.light.tc-map) — an offset headline over a horizontal row of pillar cards (Montagu Slab titles, Corbel captions) with a spotlight hover. Supports a "Core layer" always-visible block and an "Operational layer" click-revealed block, plus an optional .tc-sync indicator that reveals after the step.
6. MCP hub (.slide.light.mcp) — the "USB-C for agents" slide. Big eyebrow + acronym expansion (.mcp-acronym.mcp-expand.step.typed), three feature boxes (.mcp-box), and a chip cluster with SVG spokes (.mcp-chip + .mcp-spokes). Speedlines streak in the background.
7. Day-flow timeline (.slide.light.df-day) — a horizontal spine with numbered stops, badges that reveal on click, a pulsing "live" step and a .df-pr / .df-deploy highlight, plus an optional .df-footnote step. The full agent dev flow lives here.
8. Generic grid + chips — an offset headline over a 2×3 or 3-column grid of white cards (Montagu Slab titles, Corbel captions). Example chips can reveal on click.

Colour-panel patterns (.slide.fill — the "moment" slides; white text, yellow accent)

9. Indigo panel (.slide.fill, default) — eyebrow + headline + lead with the .mesh torus in the corner and a .wave.hero glitch wave. The brand's saturated energy.
10. Magenta / coral panel (.slide.fill.magenta, .slide.fill.coral) — colour modifiers. Use coral for chapter closers, magenta for accents.
11. Stage tracker (.htrack + .htrack-cap) — a horizontal row of .htrack-box chips with three states: .done (greyed), .on (lit indigo with .hb-dot), .todo (faint). Sits inside an indigo .slide.fill and shows progress through the 9-box harness across the deck. Pair with a .htrack-cap caption.
12. Animated bridge (.slide.fill.ink + .bridge-stage) — "Fit the harness" moment. Two .bridge-pillars rise, then .planks drop in sequence, a .bridge-shimmer sweeps, and the deck breathes. Use sparingly — it's the visual climax.
13. Agent-team grid (.slide.fill.agents-team) — rows of .agent-cards with connector .team-arrows and a .team-footnote. Optionally end with a .bridge-mini breadcrumb (Knowledge → Capabilities → Orchestration).
14. Playbook bridge / pb-hero (.slide.fill.pb-hero or .slide.fill.ink.pb-hero) — "X over here is delivered, via Z, to Y over there." Left bank of tiles (.pb-tile + .pb-bank-label) → glowing indigo .pb-bridge with .pillar, .cable, .pb-deck-planks.step, and a .pb-deck-label → VS Code-style .pb-vscode window on the right. Click expands the deck and drops in labelled planks. Hover scales individual planks.
15. Demo gate (.slide.tc-demo) — black, centered. Big "Demo N" typewriter word + a small .tc-demo-launch sub-label. Used as the transition into a live demo.
16. Closing question (.slide.fill.coral.q-slide) — giant centred "Questions?" with the accent on the punctuation.

Dark-canvas content patterns (.slide.fill or dark variants)

9. Indigo panel (.slide.fill, default) — eyebrow + headline + lead with the .mesh torus in the corner and a .wave.hero glitch wave. The brand's saturated energy.
10. Magenta / coral panel (.slide.fill.magenta, .slide.fill.coral) — colour modifiers. Use coral for chapter closers, magenta for accents.
11. Stage tracker (.htrack + .htrack-cap) — a horizontal row of .htrack-box chips with three states: .done (greyed), .on (lit indigo with .hb-dot), .todo (faint). Sits inside an indigo .slide.fill and shows progress through the 9-box harness across the deck. Pair with a .htrack-cap caption.
12. Animated bridge (.slide.fill.ink + .bridge-stage) — "Fit the harness" moment. Two .bridge-pillars rise, then .planks drop in sequence, a .bridge-shimmer sweeps, and the deck breathes. Use sparingly — it's the visual climax.
13. Agent-team grid (.slide.fill.agents-team) — rows of .agent-cards with connector .team-arrows and a .team-footnote. Optionally end with a .bridge-mini breadcrumb (Knowledge → Capabilities → Orchestration).
14. Playbook bridge / pb-hero (.slide.fill.pb-hero or .slide.fill.ink.pb-hero) — "X over here is delivered, via Z, to Y over there." Left bank of tiles (.pb-tile + .pb-bank-label) → glowing indigo .pb-bridge with .pillar, .cable, .pb-deck-planks.step, and a .pb-deck-label → VS Code-style .pb-vscode window on the right. Click expands the deck and drops in labelled planks. Hover scales individual planks.
15. Demo gate (.slide.tc-demo) — black, centered. Big "Demo N" typewriter word + a small .tc-demo-launch sub-label. Used as the transition into a live demo.
16. Closing question (.slide.fill.coral.q-slide) — giant centred "Questions?" with the accent on the punctuation.
17. Three-node storyline (.cr-stage) — three-column grid with icon nodes that light up on step-reveal. .cr-node.shown .cr-icon gets an indigo glow; the centre node (.cr-n2) pulses. Rule chips stagger in with animation-delay per :nth-child. Use for code-review, CI, or any "before → engine → after" narrative.
18. Documentation storyline — same three-node concept as CR but dark-themed with a side list. .doc-stage grid with doc-flow left and doc-list right. List items animate in via :has(.doc-n2.shown).
19. Pipeline analyzer (.pa-stage) — terminal-emulator left pane | bridge connector | chat-bubble right pane. .pa-pane.left is monospace dark; .pa-pane.right is white. .pa-line.err for red error highlights; .red-word::after for inline error backgrounds. The .pb-disc bridge pulses between panes.
20. CSP portal hub-and-spokes (.csp-hub) — central radial-gradient hub (220×220 circle) with 6 absolutely-positioned .csp-cap capability cards and animated dashed .csp-spokes SVG lines. Each cap has top/left positioning around the hub.
21. Hands-up question cards (.hands-stack + .hand-card) — dark glassmorphism cards for audience engagement. Each card: emoji .hand-ico (60px circle) + .hand-q question text. Use with .step per card for click-reveal. Great for "raise your hand if…" moments.

Lavender enders

22. Q&A (.slide.lavender.qa-slide) — lavender panel #C8C3F5, a huge centred "Q&A" (.bigword.qa) with an "Any Questions?" condensed sub-label, scattered .scatter repeats (filled + .outline) bleeding off the edges, a .roster.tr torus top-right + warped .roster.bc plane bottom.
23. Thank-You (.slide.lavender.ty-slide) — same lavender + scatter + roster treatment, big "Thank / You" stacked with the second word right-aligned (.ty1 / .ty2).

Panel modifiers: .slide.fill is indigo by default; add .magenta, .coral, or .ink to vary; .slide.lavender is the light-purple #C8C3F5 word-slide (black type). The wave auto-picks a palette that reads on each background. The lavender enders use the real wireframe rosters (.roster.tr torus top-right, .roster.bc warped plane bottom) — a more visible placement of the same mesh symbols.

Slide rhythm

Light-dominant with colour punctuation. Open on the title-photo or animated hero, drop a colour panel at each chapter break / big moment (htrack stage opener, team, bridge, closing), and keep the working content on light. Don't run two colour panels back to back, and don't go more than ~4 light slides without a colour beat. The panels are where the brand's saturated energy lives.

Sequencing tip used by the working deck: pair every stage opener (a .slide.fill with eyebrow + .htrack) with a few light working slides, then a .tc-demo gate before a live demo. Close on .q-slide → .qa-slide → .ty-slide.

The signature motif — scan-line "glitch wave" + wireframe mesh

This is what makes a slide read as BC TechDays. Two reusable elements:

• .wave — a stack of coloured horizontal scan-lines whose widths trace a rough envelope (a frequency-spectrum / waveform look), skewed slightly, with an occasional horizontal "glitch" jitter. A JS generator fills every .wave on load. Tune it with:
  • data-wave-lines="48" — line count.
  • data-wave-colors="--indigo,--indigo,--magenta,--coral" — CSS-var palette, comma-separated (repeat a colour to weight it). Default is indigo-weighted.
  • Placement classes: .wave.hero (large, bottom-right corner, for title/panel/closing) and .wave.edge (vertical-centre right edge, for quieter content slides).
• .mesh — the brand's real wireframe "rosters": the warped grid plane (#bc-mesh-plane) and the torus ring (#bc-mesh-torus), taken straight from the .pptx and inlined as SVG <symbol>s. A .mesh div holds <svg><use href="#bc-mesh-torus"/></svg> (use the plane on the hero, the torus on panels). Strokes are currentColor, so .mesh-svg recolours faint ink on light slides and faint white on .slide.fill automatically. Source files also live in assets/bc-mesh-plane.svg / assets/bc-mesh-torus.svg.

Use the wave on hero, panel, and closing slides; use it sparingly (one per slide). On content-heavy slides, prefer the quieter .wave.edge or no motif at all.

About .hero-bg / .glow1-3 — the engine ships the iFacto-style dark hero with floating coloured orbs and a grid overlay, but the BC TechDays brand override (hero-bg{background:transparent !important} + hero-bg::before,::after{display:none !important}) neutralises the dark canvas + grid. The orbs themselves still render where placed, so you can use them on the "Welcome" hero to add subtle floating colour — but most slides use .wave + .mesh instead.

Logo — the real torn-ticket lockup

The real BC TechDays logo (the official .pptx artwork: torn-ticket / speech-bubble tag, "BC/NAV TECH DAYS 2026" + "mibuso.com") is inlined once as an SVG <symbol id="bc-logo"> and referenced on every slide via <a class="brandtag"><svg><use href="#bc-logo"/></svg></a>, top-right (matching the official title slide; keeps the left clear for the headline). Its fills are currentColor, so it's ink on light slides and white on colour panels automatically. The source file is assets/bc-logo.svg. Use a <div class="topnav"><span>…</span></div> next to it for the session label (e.g. "Session 02"). (Do not hand-recreate the logo — use the real symbol.)

Buttons & links

<a class="pillbtn solid" href="#">Register</a>     <!-- indigo fill, white text, pill rounded -->
<a class="pillbtn" href="#">Programme</a>           <!-- indigo outline pill -->
<a class="arrow" href="#">More<span class="chev">›</span></a>   <!-- red chevron link, indigo on panels -->
<a class="example-link" href="#">Open the PR <span class="chev">›</span></a>   <!-- bottom-left, for live demos -->

On colour panels these invert to white automatically. Example / live-demo link: drop a single .example-link (bottom-left, above the pager) when presenting alongside something clickable (a real PR, demo URL, dashboard). Real URL only in a delivered deck — href="#" is fine in the template to show the pattern, but sweep for it and fill in or strip before shipping.

Interactive: hover spotlight & image lightbox

Two opt-in patterns let the deck feel alive when the presenter mouses around. Both are wired in template.html (CSS + JS); add the class, the behaviour follows.

• .zoomable — opt-in on any card/box element. On hover the card lifts (translateY(-4px) scale(1.04)), borders go indigo, and an inset glow fires. On colour panels (.slide.fill) the spotlight flips to a white-tint instead. Use it for grids of choices, agent cards, term boxes, "stacks" rows — anywhere the audience might wonder which one you're emphasising.
• .zoomable-img — opt-in on any <img>. Cursor turns to zoom-in, hover gently scales, and click opens a full-screen lightbox (dark blurred backdrop, image centred, "esc" close button top-right). Click backdrop or press Esc to close. The overlay is built once on load and reused for every image. Use it for screenshots, diagrams, photos — anywhere the slide thumbnail is too small to read at the back of the room. You can also open the lightbox from a non-<img> element with data-lightbox="path/to/image.png" (and optional data-lightbox-alt="…") — handy for a button or caption that should pop a full-size image.

<!-- Hover spotlight on a card -->
<div class="card zoomable">
  <h3>One option</h3>
  <p>…</p>
</div>

<!-- Click-to-focus an image -->
<img class="zoomable-img" src="assets/screenshot.png" alt="Agent flow diagram" />

Don't apply .zoomable to elements that already animate on hover (like .tc-pillar or .agent-card.anchor) — it'll double up. And don't put .zoomable-img on tiny icons; reserve it for content images the audience actually needs to read.

Extended chip types

Beyond the basic grid cards, the template carries several specialised chip/card components:

• .spec-chip — Specialist/persona chip with an amber dot indicator. Use for named agent roles or people. The dot is ::before { background: var(--amber) }.
• .model-chip — LLM model chip with colour-coded dots: .opus = magenta, .sonnet = indigo, .gpt = cyan. Use for listing available models.
• .kn-chip — Knowledge-source chip with a CSS-drawn three-line stack icon (.kn-icon). Use for knowledge bases, documentation references, data sources.
• .skill-card — Skill/capability card with a CSS-drawn file icon (.skc-icon). Larger than chips; supports a title + short caption. Use for listing agent skills or extension points.

All chips share the same hover recipe: translateY(-2px) scale(1.12), indigo border, inset glow. On .slide.fill the spotlight flips to white-tint automatically.

View Transitions / element morph

The deck supports the View Transitions API for shared-element morphing between slides. When two slides both contain elements with a matching data-morph token, the transition animates the element smoothly from its position on slide N to its position on slide N+1.

<!-- Slide N: a small chip -->
<span data-morph="bcci" class="chip">BC Code Intelligence</span>

<!-- Slide N+1: expanded hero -->
<div data-morph="bcci" class="bcci-hero">
  <h2>BC Code Intelligence</h2>
  <p class="lead">…</p>
</div>

How it works: On show(n), the JS collects [data-morph] tokens from both the leaving and entering slides. If any token appears in both AND document.startViewTransition exists, it wraps the swap in a View Transition. Falls back to instant swap on unsupported browsers.

CSS: Each morphed element auto-gets view-transition-name: <token>-token. The template provides:

::view-transition-group(*-token){ animation-duration:2000ms; animation-timing-function:cubic-bezier(.2,.8,.25,1); }
::view-transition-old(*-token),::view-transition-new(*-token){ animation-duration:1400ms; }

Use sparingly — one morph per slide transition max. Best for "chip expands into hero" or "card flies to a new position" moments.

BCCI hero shrink-on-step

A hero element that smoothly transitions its size when a later step is revealed, using :has():

.bcci-hero { transition: font-size 800ms, padding 800ms, border-radius 800ms, gap 800ms; }
.slide.bcci:has(.bcci-stack-area.shown) .bcci-hero {
  font-size: clamp(26px,3vw,48px); padding: 14px 36px; border-radius: 80px;
}

Unshown steps collapse via max-height:0; overflow:hidden; margin-top:-28px so the hero stays centred until content appears.

Replay overlay (iframe-based chat replay)

Embed and play external HTML replays (from the chat-replayer skill) directly inside the deck without leaving the presentation.

Trigger: Any element with data-replay="path/to/replay.html" and optional data-replay-title="Label". Use the .replay-link class for the standard styled button:

<a class="replay-link reveal" data-replay="replays/demo.html" data-replay-title="Full loop demo">
  <span class="rl-ico">&#9654;</span> Play replay <span class="chev">&rsaquo;</span>
</a>

Overlay structure (add once at end of <body>):

<div class="replay-overlay" id="replayOverlay" aria-hidden="true">
  <div class="replay-frame-bar">
    <span class="rt" id="replayTitle">Chat replay</span>
    <span><a id="replayOpen" href="about:blank" target="_blank" rel="noopener">Open in new tab &rsaquo;</a>
      &nbsp;&nbsp;<button class="replay-close" id="replayClose">Close &times; (Esc)</button></span>
  </div>
  <div class="replay-stage"><iframe id="replayFrame" title="Chat replay" src="about:blank"></iframe><div class="replay-sheen" aria-hidden="true"></div></div>
</div>

Behaviour:

• Click [data-replay] → sets iframe src, opens overlay with scale-up animation
• Injects Esc listener into iframe content document for seamless close
• Blocks ALL deck navigation keys while overlay is open (Space, arrows, Page)
• Supports window.postMessage('bctb-replay-close') from inside the iframe
• Decorative: rotating conic-gradient border glow + sheen sweep

The .replay-link also doubles as a regular link (href + target="_blank") so it works if JS is disabled or for right-click → open in new tab.

SVG zoom overlay

Click any .zoomable-svg to open its SVG full-viewport in a dark overlay. Built dynamically — just add the class to any inline <svg>:

<svg class="zoomable-svg" viewBox="…">…</svg>

JS clones the SVG's outerHTML into a .svg-zoom-stage div. Esc or backdrop click closes.

Dogsled trail — persistent top progress tracker

A thin always-on progress tracker pinned to the top of the deck: a sled glides along a track with named stops, marking which "stage" of the talk you're in. It's the running-progress companion to the per-section .htrack stage opener (the trail is global chrome; .htrack is a one-slide chapter beat).

Markup (one #trail block just inside <body>, after #progress):

<div class="trail" id="trail" aria-hidden="true">
  <div class="trail-track">
    <div class="trail-fill" id="trailFill"></div>
    <div class="trail-sled" id="trailSled">&#128759;</div>
    <div class="trail-stops">
      <div class="trail-stop" data-leg="1"><div class="trail-dot"></div><div class="trail-label">Stage 1</div></div>
      … one .trail-stop per stage …
    </div>
  </div>
</div>

Per-slide opt-in: a slide joins the trail with data-leg="N". TRAIL_POS (in the JS) maps each leg number to a percent along the track ({'0':2,'1':10,'2':30,'3':50,'4':70,'5':90,'6':100} by default — edit it to match your stage count). On each show(), updateTrail():

• hides the trail on slides with no data-leg (and on the data-anchor="overview" slide);
• glides #trailFill / #trailSled to the leg's percent and marks stops .done / .active;
• on a stage-shift slide (data-stage set) adds .stage-zoom so the active stop zooms to spotlight the new stage;
• flips to a light treatment (.on-dark) automatically on colour panels / dark slides.

Overview slide + build-trail morph. A .slide with data-anchor="overview" holds a big inline .build-trail (the same stops, large, each .bt-stop carrying data-goto="N" to jump to the slide whose data-stage="N"). Leaving the overview slide forward to a data-leg slide, the big build-trail flies up and becomes the top trail via a shared legtrail View Transition (~2s glide). A floating #ovLink ("Overview") button appears while you're on the stage legs (STAGE_LEGS set) and jumps back to the overview. Overview cards / build-trail stops use [data-goto]; the engine wires the jumps automatically.

The template demos the whole thing: an overview slide + five data-leg/data-stage stage slides (each also showcasing a different new component).

print-pdf export mode

Add ?print-pdf to the URL and the deck lays every slide out as a static 1280×720 block, one per page, with all .step reveals forced visible and nav / progress / trail / animations stripped — then Print → Save as PDF (the @page size is set to match). Triggered by a one-line guard at the top of the main script that adds .print-pdf to <html>; the rest is the html.print-pdf … CSS block. Use it to hand out or archive a deck.

Vocabulary cards — click-to-zoom

A row of term cards that blow up to ~200% on click — a richer replacement for the old one-word-per-slide vocabulary walk. .vocab-grid holds .vocab-cards (.vc-kicker / .vc-word / .vc-desc). Clicking a card clones its content into a centred .vocab-zoom overlay (.vz-card) scaled up; backdrop click or Esc closes. The overlay is built once on load.

<div class="vocab-grid">
  <div class="vocab-card"><span class="vc-kicker">Noun</span><span class="vc-word">Harness</span><span class="vc-desc">Short definition, zoomed on click.</span></div>
  …
</div>

More layout components

Pure-CSS content blocks added to the template; copy the markup, swap the content.

• .term-grid — a row of .term-item cards (indigo top-rule) with a Barlow-condensed .term-label + .term-def. A terminology / glossary strip on light slides.
• .strap-list — vertical numbered rows (.strap-item → .strap-num yellow + .strap-text + optional right-aligned .strap-pillar). Takeaways / principles on a colour panel.
• .photo-split — text left (.ps-text), framed image right (.ps-img); the image takes .zoomable-img or data-lightbox for click-to-zoom. .instr-split is the same idea with a narrower image column (.is-text / .is-img) for screenshots.
• .history-gallery — an equal-width row of photos (dark slides), each lightly dimmed; pairs with .zoomable-img.
• .fail-grid — a 2×2 grid of captioned thumbnails (.fail-img-wrap + .fail-img-caption overlay). "Where it breaks" / before-and-after evidence.
• Retro C/AL skin (.slide.retro-cal + .crt) — a deliberately old-fashioned Win95 / C-SIDE window (.crt-title / .crt-menu / .crt-body with .crt-eyebrow + .crt-h2 + .crt-p, .crt-status footer) on a teal desktop. The "we used to do it the OG way" gag slide. The wave/mesh motif is hidden here on purpose.
• .replay-row — several .replay-link launch buttons side by side in flow (centred), instead of one bottom-left button. Use when a slide offers a menu of replays to play in-deck.

Presenter notes (inline in the slide picker)

Per-slide talk-track notes accessible from the slide picker, for speaker preparation and on-stage reference. (This replaced the old .narr modal: notes now expand inline under the picker row — no separate overlay.)

Define notes in a SLIDE_NOTES object keyed by data-slide — one plain string per slide:

const SLIDE_NOTES = {
  '00': 'Title. Welcome them in. 45 minutes, two live demos.',
  '02': 'Walk the prerequisites — keep it to one sentence out loud.',
};

UI: When building the slide picker, each slide that has a SLIDE_NOTES entry grows a small ⓘ button (.picker-note-btn). Clicking it toggles an expanding .picker-note row right under that picker item; the picker stays open. If SLIDE_NOTES is undefined or has no entry for a slide, no ⓘ appears. Keep each note to a sentence or two — it's a cue, not a script.

Slide persistence

Two strategies wired in the deck engine — use one or both:

• URL hash (default): history.replaceState(null,'','#'+(i+1)) on every slide change. Reloading the page resumes at the same slide. Shareable links (deck.html#14). Responds to hashchange for browser back/forward.
• localStorage (opt-in): localStorage.setItem('bctd-current-slide', i) — survives reloads even without hash support. Useful for offline decks or when you don't want the hash visible.

tc-sync pattern (:has() sibling reveal)

A row that auto-reveals when a sibling .step gets .shown, without needing its own click:

.slide.tc-map .tc-map-row.tc-sync { opacity:0; transform:translateY(28px); }
.slide.tc-map.active:has(.step.shown) .tc-map-row.tc-sync {
  animation: revealUp .8s cubic-bezier(.2,.8,.25,1) forwards;
}

Use for "appears together with the first step" elements — sync indicators, footnotes, secondary labels.

animation-fill-mode: backwards technique

Use backwards instead of forwards when you need :hover transforms to override the animation's final state:

.slide.mcp.active .mcp-boxes.step.shown .mcp-box {
  animation: mcpBoxSnap .55s cubic-bezier(.34,1.56,.64,1) backwards;
}

backwards holds the initial keyframe during animation-delay, then after completion the property reverts to the regular cascade — so :hover { transform:… } works freely.

Topbar anti-animation fix

Prevent the persistent topbar from re-animating on every slide change:

.slide .topbar.reveal { animation:none; opacity:1; transform:none; }

Step reveal, typewriter & present mode

• Add class step to any element to hide it until the next Next-click / Right-arrow / Space. Combine step + typed for a typewriter reveal (per-char animation; the JS at the bottom splits text into per-char spans on load). Steps reset when the slide changes.
• An element with .typed but no .step types automatically when the slide becomes active (250ms initial pause then ~22ms/char).
• ?present=1 in the URL (or pressing P) hides the bottom nav for a clean projected view.
• Click the slide counter to open the slide picker (jump to any slide); the picker is grouped by data-topic via .picker-topic-sep headers, with .picker-item.current highlighting the active slide. Esc closes it.

Things to get right — don't repeat the easy mistakes

• The presentation brand is light canvas + Montagu Slab + RED titles, not dark + lime, and the title colour is red #F7514F, not purple/indigo. (The marketing site's CSS exposes a lime token and indigo reads as purple — neither is the title colour. The .pptx "main title" is red.)
• Headlines are Montagu Slab (slab serif). Body is Corbel (with a Hanken Grotesk fallback). Eyebrows/numbers are Barlow Condensed. The h1-h4 override carries !important — don't fight it with inline font-family unless you have a reason.
• Use the real artwork, don't recreate it: the logo (#bc-logo) and the wireframe rosters (#bc-mesh-plane / #bc-mesh-torus) are inlined from the official .pptx. A hand-drawn logo was a rejected earlier attempt.
• Indigo is the panel/component colour; red is the title colour. The other colours are for categorical meaning — don't scatter all of them on one slide.
• The .g accent word is ink on light, yellow on colour panels. The template flips it automatically; don't override it back.
• Use the glitch-wave / roster-mesh motif to brand a slide — but one motif per slide, keep it in a corner clear of text, and keep content slides quiet.
• One stage opener per stage: don't spam .htrack on every slide — it's a chapter beat, not chrome.
• The dogsled trail is the global progress chrome — only tag the slides that belong to a stage with data-leg (and data-stage on the stage-shift slide). Slides with no data-leg hide it. Keep TRAIL_POS and the .trail-stop count in sync with your stage count, and put the .build-trail on the data-anchor="overview" slide so the morph has something to fly from.
• The animated bridge and pb-hero are heavy hitters; one per deck (each). The build-trail → top-trail morph is the same kind of one-shot moment — one overview slide, once.
• No dead links in a delivered deck. Sweep for href="#" and placeholder data-replay/data-lightbox/src paths (the template's demo images use assets/title-bg*.jpg).
• Before printing/archiving, open with ?print-pdf and "Save as PDF" rather than screenshotting slides.
• Default language is English; translate placeholder copy to the real talk.

Output

Save to the user's workspace folder. Give them a computer:// link so they can open it. Don't put the deliverable in this skill folder.

Reference

• Brand source: the official BCTechDays2026 PowerPoint (theme colours, embedded fonts: Montagu Slab + Corbel; logo/motif SVGs). Re-open it with the pptx skill to re-verify a token.
• Live site: https://www.bctechdays.com (and mibuso.com). Treat its raw CSS tokens with caution — verify against the deck artwork before trusting a colour.
• Shares the deck engine with the ifacto-presentation skill — keyboard nav, picker, typewriter, wave generator, present mode all behave identically.