← 返回 Skills 市场
mupengi-bot

Cardnews

作者 mupengi-bot · GitHub ↗ · v1.0.0
cross-platform ⚠ suspicious
662
总下载
0
收藏
2
当前安装
1
版本数
在 OpenClaw 中安装
/install cardnews
功能描述
Generate Instagram-ready card news (카드뉴스) image sets. Use when creating a series of 5 slide images from a topic — includes content planning, image generation...
使用说明 (SKILL.md)

Card News (카드뉴스) Pipeline 🐧

Topic → 5-slide plan → image generation → JPG conversion → caption → Instagram upload.

Workflow

1. Content Planning

Given a topic, plan 5 slides:

Slide Role Content
1 Hook Bold question or surprising statement to stop scrolling
2 Problem/Context Why this matters
3 Core insight Key information or explanation
4 Detail/Example Supporting evidence or practical example
5 CTA/Summary Takeaway + follow/save prompt + 🐧 branding

2. Image Generation (nano-banana-pro)

Generate each slide as 1024×1024 PNG using the nano-banana-pro skill:

# Path may vary based on installation
uv run \x3Copenclaw-install-dir>/skills/nano-banana-pro/scripts/generate_image.py \
  --prompt "\x3Cslide prompt>" --filename "cardnews-TOPIC-N.png" --resolution 1K

Prompt structure per slide:

  • Include exact Korean text to render (in quotes)
  • Specify: dark background, neon cyan (#00FFFF) accent text, white primary text
  • Include 🐧 penguin emoji on slide 5
  • No English unless the topic requires it

See references/design-guide.md for visual style rules.

3. PNG → JPG Conversion

Instagram rejects PNG frequently. Always convert:

python3 {baseDir}/scripts/convert_jpg.py cardnews-topic-1.png cardnews-topic-2.png ...

Outputs *-ig.jpg files (1024×1024, quality 92).

4. Caption Writing

Write an Instagram caption in Korean:

  • 1-2 line hook matching slide 1
  • 3-5 hashtags (mix of broad + niche)
  • End with CTA: "저장하고 나중에 다시 보세요 📌"

5. Instagram Upload

Use the browser-based upload flow documented in TOOLS.md ("인스타그램 게시물 업로드" section). Upload all 5 JPG files as a carousel post.


🐧 Built by 무펭이무펭이즘(Mupengism) 생태계 스킬

安全使用建议
This skill appears coherent and small. Things to consider before installing: (1) It relies on another skill (nano-banana-pro) — verify that skill's source and permissions because image generation is delegated to it. (2) The included convert_jpg.py will pip-install Pillow at runtime if missing, which downloads packages from PyPI — if you require an offline or locked environment, pre-install Pillow or inspect and sandbox the execution. (3) The SKILL.md references a TOOLS.md browser upload flow not included here — confirm how uploads are handled in your environment. If you plan to run this in a shared or sensitive environment, run the converter in a sandbox or review/approve the nano-banana-pro skill first.
功能分析
Type: OpenClaw Skill Name: cardnews Version: 1.0.0 The `SKILL.md` instructs the OpenClaw agent to construct and execute shell commands (`uv run`, `python3`) where parameters like image prompts and filenames are derived from user input (the 'topic'). This creates a significant shell injection vulnerability if the agent does not properly sanitize or escape user-provided input before executing these commands, potentially leading to arbitrary command execution. While the `scripts/convert_jpg.py` file itself is benign, its execution is part of this vulnerable command construction pattern.
能力评估
Purpose & Capability
Name/description match the content: planning slides, generating images via another image-gen skill (nano-banana-pro), converting PNG→JPG, caption writing, and upload prep. Required binaries/env/config are empty and consistent with an instruction-only skill that delegates image creation to another skill.
Instruction Scope
SKILL.md stays on-task (plan slides, call nano-banana-pro to generate PNGs, convert to JPG, write caption, use browser upload). It references running a script from the nano-banana-pro skill and a browser-based TOOLS.md workflow — these are reasonable but require the nano-banana-pro skill and the TOOLS.md upload doc to be present and trusted. The instructions don't ask for unrelated files, credentials, or system-wide config.
Install Mechanism
No install spec (instruction-only) which is low-risk. The included convert_jpg.py will attempt to pip-install Pillow at runtime if it's missing (via subprocess.check_call). That behavior performs a network package install when first run — not inherently malicious but worth noting because it writes packages to the environment at runtime.
Credentials
Skill declares no environment variables, no credentials, and no config paths. SKILL.md does not reference environment variables or secrets. The lack of external credential requests is proportionate to the stated purpose, though the skill delegates image generation to another skill which may have its own credential needs.
Persistence & Privilege
always is false and the skill does not request permanent presence or modify other skills. It contains a small utility script but does not change system or agent-wide settings.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install cardnews
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /cardnews 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
- Initial release of the cardnews skill for generating Instagram-ready 카드뉴스 image sets. - Automates the full pipeline: topic-to-content planning, 5-slide image generation (using nano-banana-pro), PNG→JPG conversion, caption creation, and upload prep. - Ensures design consistency (dark background, neon cyan and white text, Korean language focus). - Delivers complete Instagram carousel post assets and workflow guidance.
元数据
Slug cardnews
版本 1.0.0
许可证
累计安装 2
当前安装数 2
历史版本数 1
常见问题

Cardnews 是什么?

Generate Instagram-ready card news (카드뉴스) image sets. Use when creating a series of 5 slide images from a topic — includes content planning, image generation... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 662 次。

如何安装 Cardnews?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install cardnews」即可一键安装,无需额外配置。

Cardnews 是免费的吗?

是的,Cardnews 完全免费(开源免费),可自由下载、安装和使用。

Cardnews 支持哪些平台?

Cardnews 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Cardnews?

由 mupengi-bot(@mupengi-bot)开发并维护,当前版本 v1.0.0。

💬 留言讨论