gpt-image-slide
End-to-end GPT image slide workflow. Trigger when the user wants to create a slide deck as generated images from a reference slide, source files, and a deck prompt. Runs design, plan, prompt, and render stages in order.
How do I install this agent skill?
npx skills add https://github.com/humantonylee/future-slide --skill gpt-image-slideIs this agent skill safe to install?
- Gen Agent Trust Hubpass
The skill orchestrates a multi-step workflow to generate slide images from user-provided files and reference images. No malicious code, obfuscation, or high-risk command execution was detected. A low-severity risk of indirect prompt injection exists because the skill processes untrusted user data to generate prompts and images without explicit sanitization or boundary markers.
- Socketpass
No alerts
- Snykpass
Risk: LOW · No issues
What does this agent skill do?
GPT Image Slide Workflow
Use this skill when the user asks for the full GPT-image slide workflow, for example:
$gpt-image-slide- "create slide images from this reference slide and report"
- "make a deck with GPT image generation"
- "run the full future-slide image workflow"
This is the orchestrator. If the user asks for only one step, use the matching step skill instead:
$slide-designforDESIGN.md$gpt-image-slide-planforslide_plan.json$gpt-image-slide-promptforslide_prompts.json$gpt-image-slide-renderfor finalpage_<n>.pngimages
Required Sequence
Run these stages in order:
- Extract
DESIGN.mdfrom the reference slide image or visual reference. - Build
slide_plan.jsonfromDESIGN.md, user files, and the deck request. - Build
slide_prompts.jsonfromDESIGN.mdandslide_plan.json. - Render page images sequentially with Codex native image generation.
Do not skip ahead. Do not render before the prompt JSON exists.
Output Contract
The full workflow should leave these files in the workspace:
DESIGN.mdslide_plan.jsonslide_prompts.jsonpage_1.pngthroughpage_N.png
Routing Rule
If the user mentions HTML, browser-rendered slides, editable web slides, or export from HTML, use the html-slide skill family instead.
How can the creator link this skill?
Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.
<a href="https://skillzs.dev/skills/humantonylee/future-slide/gpt-image-slide">View gpt-image-slide on skillZs</a>