dispatch
Shared multi-model CLI dispatch infrastructure for the adv plugin. Houses dispatch.sh, preflight.sh, run-phase.sh, scope.sh, and reviewer prompt templates used…
Create visually rich, style-driven slide decks in two output formats: PowerPoint (.pptx via pptxgenjs + React icons) and HTML (single-file Reveal.js deck). Use any time the user asks for a slide deck, presentation, pitch deck, training module, keynote, .pptx file, HTML/web
$ npx -y skills add andisab/swe-marketplace --skill slideware --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/slidewareContext preview
The summary Claude sees to decide when to auto-load this skill.
Create visually rich, style-driven slide decks in two output formats: PowerPoint (.pptx via pptxgenjs + React icons) and HTML (single-file Reveal.js deck). Use any time the user asks for a slide deck, presentation, pitch deck, training module, keynote, .pptx file, HTML/web
name: slideware description: "Create visually rich, style-driven slide decks in two output formats: PowerPoint (.pptx via pptxgenjs + React icons) and HTML (single-file Reveal.js deck). Use any time the user asks for a slide deck, presentation, pitch deck, training module, keynote, .pptx file, HTML/web presentation, Reveal.js deck, browser slides, or slides to share via link — or wants slides generated from a brief/outline, references a 'style' or 'brandbook' (a .md/.css visual identity), a 'sample deck' to mimic, or asks to update/iterate on an existing deck (rebuild approach). Trigger on words like 'deck', 'slides', 'presentation', 'pptx', 'pitch', 'keynote', 'reveal', 'HTML deck', 'web presentation', or whenever a .pptx (or slide-deck .html) appears as input or output. Styles resolve from the knowledgeware plugin's shared registry — 5 generic defaults plus brandbooks (in styles/brands/ or the user's KNOWLEDGEWARE_BRANDS_DIR) managed by the brandware skill — so decks share one visual identity with knowledgebase sites and chartware diagrams. If the output format (PowerPoint vs HTML) is not stated and cannot be inferred from context, ASK the user before building." license: MIT
Generate brandbook-driven slide decks in one of **two output formats**:
| Format | Engine | Output | Guide | |---|---|---|---| | **pptx** | pptxgenjs + React icons + sharp (Node.js) | Real `.pptx` — opens in PowerPoint, Keynote, Google Slides | [pptx/FORMAT.md](pptx/FORMAT.md) | | **html** | Reveal.js + js-yaml (Node.js) | Single self-contained `index.html` — any browser, hostable, share-by-link | [html/FORMAT.md](html/FORMAT.md) |
Opinionated about one thing in both formats: **don't make boring slides.** Plain bullets on white aren't worth your reader's time.
Decide from explicit cues first:
**If neither the request nor context settles it, ASK the user** ("PowerPoint file or an HTML deck you can open in a browser/host?") before building — the two outputs are not interchangeable deliverables. Don't silently default.
Then read **exactly one** format guide (`pptx/FORMAT.md` or `html/FORMAT.md`) and follow it. Each guide has its own workflow, layout archetypes, pitfalls, and scripts under its directory.
"Style" and "brandbook" mean the same thing — .md files describing a visual identity (palette + type + layout). Path convention: `$PLUGIN` = the knowledgeware plugin root, two directories above this skill (`../../`).
1. **Explicit local path** — read directly. 2. **Bundled name** — discovered at runtime from the registry (shadow order: default < plugin brand < user brand):
Run `node $PLUGIN/scripts/list-styles.js` for the current list (`--default` / `--brands` / `--names` / `--json`). 3. **Website URL** — `node $PLUGIN/scripts/derive-style.js <url> -o ./brand.md` then load. Heuristic; verify against the live site. 4. **Google Drive URL** — `bash $PLUGIN/scripts/fetch-resource.sh <url> <dest>`. 5. **No style given** — if a `DEFAULT` marker exists (a one-line file naming a registry entry; `$KNOWLEDGEWARE_BRANDS_DIR/DEFAULT` wins over `$PLUGIN/styles/brands/DEFAULT`), use that brand and tell the user; otherwise pick `style-1` (Editorial Light) and tell the user. Don't otherwise fall back to brandbooks unless the user names one.
**Staging differs per format** (each guide shows the exact command):
**Fonts (handle silently)**: brandbooks name fonts; nothing embeds them. When building **pptx**, after resolving the style run `bash $PLUGIN/scripts/install-fonts.sh --check <name>` and, if Google Fonts are missing, install them with the same script without asking — just mention it in one line. Non–Google-Fonts families can't be auto-installed: fall back per the brandbook and tell the user which face to install manually. HTML decks load Google Fonts via `<link>`, so viewers (and you) need nothing installed.
**Missing tooling**: pptx previews need LibreOffice + poppler (`render-slides.sh` prints a NOTE and exits cleanly when they're absent — no PNGs means the check was skipped) and visual review needs browser MCP tools (find them via `tool_search`). When any of these are missing, still build and deliver the deck, say which check was skipped, and point the user to the plugin README's "Optional tooling" section for install commands.
The point is **content-shape variety**, not visual-archetype rigidity. A deck that walks one consistent story can reuse a layout deliberately; a survey of unrelated topics should vary archetypes slide to slide. The failure mode is reaching for the same layout because the template defaulted to it. If you reuse an archetype twice without articulating why, swap one out. Archetype catalogs: [pptx/references/layout-patterns.md](pptx/references/layout-patterns.md) (12) · [html/references/layout-patterns.md](html/
A curated Claude Code plugin marketplace for practical, everyday usage in software engineering — 13 plugins, 53 specialist agents, 14 skills, 3 commands. A few opinionated choices that set it apart from larger awesome-style lists: Curated, not exhaustive.
Repo: andisab/swe-marketplace
Shared multi-model CLI dispatch infrastructure for the adv plugin. Houses dispatch.sh, preflight.sh, run-phase.sh, scope.sh, and reviewer prompt templates used…
Use this skill when creating or refining Claude Code sub-agent definitions. Helps design specialized AI assistants with proper YAML frontmatter, system…
Use this skill when creating or refining custom Claude Code slash commands. Slash commands are user-invoked reusable prompts that can accept arguments,…
Use this skill when creating or refining Claude Code hooks. Hooks are shell commands that execute at specific lifecycle events (tool use, prompt submit,…
Create MCP servers — multi-tool services exposed via Model Context Protocol. Use this skill whenever users mention MCP servers, building servers, server…
Create MCP tools — individual tool functions exposed via Model Context Protocol. Use this skill whenever users mention MCP tools, tool handlers, tool…