Deck AI

What it does

When your agent uses it

• Turning a markdown outline into a PDF presentation
• Building a slide deck with photos from lesson or pitch notes
• Making a Gamma-style deck without a subscription

Example prompts

• “Make a deck from notes/quarterly-update.md and export it as a PDF.”
• “Turn this markdown into slides with a cover, a quote slide and image layouts.”
• “Build slides from ./lesson-plan.md with photos that match each slide's topic.”

Requirements

• Node 18 or newer and pnpm
• An Unsplash access key in a .env file
• A Slidev workspace created by the bundled setup script

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

1. 1Reads source markdown (slides separated by ---).
2. 2Picks a layout + image keyword per slide. Two modes
3. 3Fetches images from Unsplash by keyword.
4. 4Renders the whole thing to PDF via Slidev.

What it can do on your machine

Read from SKILL.md and the folder at commit 16d4724. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Ships 3 files in scripts/ (Python and Shell), which the agent can run.

  Shell commands in SKILL.md call:

  • bash
  • python3
  • npm
  • brew
  • pnpm

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  Links to these hosts (documentation or services it may open):

  • unsplash.com
  • sli.dev

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names these keys or tokens, usually read from environment variables:

  • UNSPLASH_ACCESS_KEY

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Deck AI loads about 1.5k tokens when it runs, and up to ~1.9k if it reads all its reference files. Until then it costs about 67 tokens; SKILL.md has 549 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~67

When it runs · the whole SKILL.md, loaded when a task matches

~1.5k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~1.9k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

• NoteMentions a .env fileSKILL.md:40

  - Key lives in `.env` at the calling instance root:

• NoteMentions a .env fileSKILL.md:42

  ho 'UNSPLASH_ACCESS_KEY=<your-key>' >> ./.env

• NoteMentions a .env fileSKILL.md:44

  - Never commit `.env`. Never paste the key into any file in this skill.

• NoteMentions a .env fileSKILL.md:164

  `UNSPLASH_ACCESS_KEY not set` | Add to `.env` at calling dir |

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from assafkip/kipi-system at commit 16d4724, republished under its MIT licence (© assafkip). 549 words, ~1,480 tokens.

name

deck-ai

description

Generate modern presentation decks (PDF) from markdown content. Local open-source alternative to Gamma — uses Slidev for layouts and Unsplash for imagery. Invoke when the user asks to "make a deck", "build slides from this", or "turn this into a presentation".

user-invocable

false

bash <skill>/scripts/setup.sh ./deck-workspace

# 1. Preflight: key + deps must be present
bash <skill>/scripts/check_env.sh ./deck-workspace

# 2. (LLM mode) Claude reads <input.md>, decides layout + image keyword
#    per slide, and WRITES ./deck-workspace/decisions.json with the Write
#    tool (schema below). Skip this step when running outside Claude.

# 3. Generate Slidev markdown from source
#    LLM mode:
python3 <skill>/scripts/generate.py <input.md> ./deck-workspace/slides.raw.md \
    --theme seriph --decisions ./deck-workspace/decisions.json
#    Rules mode (no decisions file):
python3 <skill>/scripts/generate.py <input.md> ./deck-workspace/slides.raw.md --theme seriph

# 4. Fill image placeholders from Unsplash
python3 <skill>/scripts/fetch_images.py ./deck-workspace/slides.raw.md

# 5. Export PDF
bash <skill>/scripts/render.sh ./deck-workspace ./deck-workspace/slides.raw.md ./deck.pdf

[
  {"layout": "cover",       "image_keyword": "sunrise mountains"},
  {"layout": "image-right", "image_keyword": "student classroom"},
  {"layout": "center",      "image_keyword": null},
  {"layout": "end",         "image_keyword": null}
]

# Deck Title

Subtitle line.

---

# Second slide

Body content here.

<!-- layout: image-right -->

Anime anchor: Naruto.

---

# Homework

Optional closing text.

cd ./deck-workspace && pnpm add @slidev/theme-<name>

© assafkip, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 4 other files (scripts, references) in plugins/kipi-core/skills/deck-ai of assafkip/kipi-system.

• SKILL.md
• references/layout-catalog.md
• scripts/check_env.sh
• scripts/fetch_images.py
• scripts/render_pptx.py

Open the folder on GitHubat commit 16d4724