CardPrinter is an open-source tool that automatically generates Instagram carousel images and Reels videos from a single topic line, with built-in quality gates and batch production support.

Stars

11

7-day growth

No data

Forks

5

Open issues

0

License

NOASSERTION

Last updated

2026-07-05

AI repository intelligence
FR-AI / ANALYSIS

Why it is worth attention

It enforces physical quality checks (contrast, overflow, fact-fidelity) and uses a self-healing factory runner for mass production, offering a free, self-hosted alternative to Canva bulk-create and Predis.ai.

Who it is for

  • Social media marketers and content creators
  • Developers building automated content pipelines
  • Small businesses or agencies needing branded carousels
  • Journalists and publishers repurposing news into posts

Use cases

  • Converting a news article into a ready-to-post Instagram carousel and Reels
  • Batch-generating brand-consistent carousels for a campaign
  • Automating social media content from a list of topics with minimal human effort
  • Producing fact-checked and WCAG-compliant visual stories

Strengths

  • Fully open-source, free, and self-hosted with no watermarks
  • Produces both carousel and Reels from the same copy, ensuring consistency
  • Hard quality gates (WCAG contrast, overflow, fact-fidelity) prevent broken output
  • Self-healing factory runner for parallel batch production with automatic retries

Considerations

  • Requires manual installation of Node.js 20+, Python 3.11+, and Chrome headless
  • Output limited to Instagram formats (4:5 carousel and 9:16 Reels) only
  • Fact-fidelity gate relies on external copywriting; may need human oversight for accuracy

README quick start

CardPrinter 🖨️

Turn one topic line into a ready-to-post Instagram carousel and a 9:16 Reels video — fully automated, quality-gated, zero design skills required.

English | 한국어

An open-source alternative to Canva bulk-create, Predis.ai, and Ocoya — but the output is measured against physical quality gates, not vibes.


Demo

Carousel (4:5)

Photo article (newsprint)Numbered roundup (newsprint)Photo feature (newsprint)

Hero above: the newsprint cover system — modeled on a 300K-like Instagram carousel.

Reels (9:16)

https://github.com/kimsh-1/cardprinter/raw/main/docs/demo-reel.mp4

Every carousel is derived from the same copy contract as its Reels cut, so the two never drift apart. (The clip above is a real pipeline output; GitHub renders docs/demo-reel.mp4 inline.)


✨ Features

  • 🗞️ One-line input — give it a topic (or a news article); it writes the copy, lays out the cards, and renders both formats.
  • 🎨 Tone tiersluxury · character · brand · news · editorial · data, plus custom tiers auto-built from a company's BI deck or a 5-question interview.
  • 📐 Viral sequence templates — checklist · listicle · data-infographic · photo-story · storytelling · versus, each a research-backed slide flow (triple-hook, mid save-trigger, end CTA).
  • 🛡️ Physical quality gates — contrast (WCAG pixel-measured), overflow (bbox ⊂ safe zone), fact-fidelity (no fabricated numbers), typography floor, layout diversity — all exit-code hard gates, not suggestions.
  • 📊 Honest charts — echarts SSR bar charts with a 0-based scale; the display string is the source of truth so numbers can't be exaggerated.
  • 🎬 Card-news → Reels — finished cards are composited into a 1080×1920 timeline with clean rise-in transitions (list/chart items reveal one by one — no zoom, no jitter), silent kinetic captions, per-card dwell derived from reading speed.
  • 🏭 Mass production — a headless codex factory runner produces topics in parallel with a self-heal loop (failed gates are fed back and regenerated).

Why this instead of Canva / Predis.ai?

CardPrinterSaaS bulk-create
PriceFree, open source

Description

Turn one topic into a gate-checked Instagram carousel + 9:16 Reels — fully automated, quality-gated, self-hostable.

Related repositories

Similar projects matched by category, topics, and programming language.

op7418
op7418 GitHub avatar

guizang-material-illustration

A skill for AI agents like Claude Code and Codex to generate explanatory illustrations with Chinese labels in the 'Guizang' material style, making abstract concepts, charts, and reference-based content into clear center visuals.

Empty
835
OneInterface
Featured
OneInterface GitHub avatar

stormy-cookbook

Stormy Cookbook provides copy-pasteable recipes for using the Stormy Social Data API and MCP server, which offers a unified REST API and MCP tools to search profiles, posts, and emails across TikTok, YouTube, Instagram, LinkedIn, X, and Reddit through a single HTTP contract.

AI & Machine LearningLarge Language Models
72
yanhua1010
yanhua1010 GitHub avatar

self-media-content-workflow

A modular, tool-agnostic suite of Agent skills for the complete self-media content production lifecycle, from fuzzy brief to delivery and analytics, with mandatory human-in-the-loop checkpoints.

Python
295