Bento Slides: Install, Source and Security | FunnelSlayer

Bento Slides

Published by shenyuannext in wepr-growth-skills

Review recommended1 installs

What this skill does

>-

Add Bento Slides to your agent

Review the source and files first. When you are ready, copy the prompt instruction or use the CLI command supported by your environment.

Install with a prompt

Paste this into a compatible coding agent:

add this skill "bento-slides" from https://github.com/shenyuannext/wepr-growth-skills

Install with the CLI

Run this command in a controlled environment after reviewing the repository:

npx skills add https://github.com/shenyuannext/wepr-growth-skills --skill bento-slides

Skill instructions

Authoring Bento decks

For WEPR proposals, strategy decks, quotation presentations, quarterly reviews, and client reports, first read references/wepr-delivery.md. Keep claims, figures, pricing, platform scope, and source dates consistent with the approved source document or spreadsheet.

A Bento deck is one self-contained .bento.html file. The document is plain JSON in a single block:

<script type="application/bento+json" id="bento-doc"> { "format":"bento/slides", ... } </script>

You edit that block only, in place. Escape every < in the JSON as \u003c so it can never contain a literal </script>. Leave the rest of the file (the compressed runtime) untouched. In a chat context instead, the user copies the JSON out (Save ▾ → Copy document JSON) and pastes your replacement back (Save ▾ → Replace from JSON…); window.bento.loadDoc(json) does it from the console.

Starting from nothing

The user does NOT need Bento installed — the app ships inside every deck. When there is no .bento.html to edit, fetch the latest signed release yourself and author into it:

# name the file after the deck's topic, e.g. Q4_Review.bento.html
curl -fsSL https://bento.page/releases/slides/Bento_Slides.bento.html -o "<Topic>.bento.html"

(Windows without curl: iwr https://bento.page/releases/slides/Bento_Slides.bento.html -OutFile <Topic>.bento.html.)

Then verify the download contains id="bento-doc", and replace that block's JSON (it ships with a showcase deck — discard it) with your document. Use python3 scripts/bento_deck.py inject <deck.bento.html> <document.json> for deterministic replacement and python3 scripts/bento_deck.py validate <deck.bento.html> before delivery. Current shells may contain an empty #bento-doc block; that is valid and ready for injection. Rules for a fresh document:

  • Fetch https://bento.page/agents.md BEFORE authoring and start from its "Minimal valid document" skeleton. size and theme (including theme.fontFamily) are required — the app will not boot without them.
  • Fully specify element fields as the skeleton shows (shapes need stroke/strokeWidth; text needs fontFamily/align/valign) — missing fields render wrong or not at all.
  • Omit docId and collab entirely: the app mints a fresh identity and dormant collaboration credentials on first open.

When done, offer to open it (open / xdg-open / start) — the file boots straight into the editor with the finished deck. Aim for one pass from request to opened deck.

Workflow

  1. Find the document. Locate the #bento-doc block; parse its JSON. Note doc.size (canonical 1280×720), doc.theme, existing element ids, and whether doc.template/doc.readonly are set.
  2. Read the source material the user gave you and classify each piece — is it a stat? a table? a process? a definition to expand? a photo?
  3. Map material → feature (do NOT default to bullet text). This is the step that makes it a Bento deck rather than a slideshow of paragraphs:
    • numbers to compare visually (trend, magnitude, share) → a chart element
    • a comparison / spec / pricing / feature grid → a table element (columns weights + rows of cells + a style object)
    • consecutive slides about the same thing changing → morph: give shared elements the same id on both slides + transition:"morph" on the later one (Bento's signature move — reach for it liberally)
    • a point to drill into → a state slide (stateOf + element link)
    • a hero / full-slide image → full-bleed image + scrim rect + text, with ken-burns drift
    • a sequence / flow / timeline → a line/path with a dash-march loop, or morph a highlight through the steps
    • a headline number → big text + fx:{countUp:true}
    • every cover / divider → at least one ambient motion
    • repeated chrome / logo → keep its id stable across slides so it morphs in place
    • a demo clip / recording / soundbite → a media element (kind: video|audio); embed short clips as a data URI, link big ones by URL to keep the file small
  4. Author using the schema. Keep the full schema and copy-paste recipes open: fetch https://bento.page/agents.md (it has the element shapes, the morph/chart/state/ken-burns snippets, and the gotchas). Respect one accent colour, ≤2 typefaces, 96px side margins (right-most x ≤ 1184), and write speaker notes on each slide.
  5. Self-audit before finishing:
    • any numbers rendered as text that should be a chart?
    • do consecutive slides on one subject share ids + transition:"morph"?
    • at least one motion moment (ken-burns / loop / count-up), esp. the cover?
    • a drill-down that would work better as a state slide?
    • one accent colour, ≤2 typefaces, 96px margins?
    • speaker notes on every slide?
  6. Write back the edited #bento-doc block (escaping <), or return the replacement JSON. Never regenerate the whole HTML file.
  7. Validate with scripts/bento_deck.py validate, then open the file in a browser and review every slide, interaction, chart, note, and margin.

Critical gotchas

  • Charts: bar/line series data must be plain numbers ({value,…} item objects coerce to 0 — only pie takes {name,value}); colour by series, not per bar; option is pure JSON, template formatters only ({b}/{c}/{d}), never functions.
  • Morph needs deterministic, stable ids shared across the slides that should animate together. Different ids = no morph (elements just cut).
  • Images/fonts must be embedded as data URIs in doc.assets and referenced by "asset:<key>" — the file stays self-contained.
  • Media: a media element (kind: video|audio) embeds short clips as a data URI in src (self-contained) or references a URL for big files (keeps the deck small). autoplay runs only in present mode and needs muted:true for video. Don't embed large videos — they bloat the file.
  • Never regenerate docId when editing an existing deck; it is the document's identity. (Fresh decks omit it — the app mints one.)
  • template:true → every open mints a fresh deck; readonly:true → the file boots straight into the show with no editor.

Working examples of every technique: open any template at https://bento.page and read its #bento-doc block.

Files included

  • agents/openai.yaml
  • LICENSE
  • references/wepr-delivery.md
  • scripts/bento_deck.py
  • SKILL.md

More skills