← Back to Skills Marketplace
Og Image Design
by
Ömer Karışman
· GitHub ↗
· v0.1.5
815
Downloads
0
Stars
3
Active Installs
2
Versions
Install in OpenClaw
/install og-image-design
Description
Open Graph and social sharing image design with platform specs, text placement, and branding. Covers OG meta tags, Twitter cards, LinkedIn previews, and dyna...
Usage Guidance
This skill appears to do what it says: generate Open Graph/social images using the inference.sh CLI. Before installing or using it: 1) Treat `curl https://cli.inference.sh | sh` as potentially risky — prefer to download the installer and verify the SHA-256 checksum from the provided checksums.txt before running. 2) Expect that example HTML and image content will be sent to inference.sh; do not send sensitive data (secrets, private user data, proprietary HTML) unless you trust their privacy policies. 3) `infsh login` implies credentials or tokens — confirm how those are stored and revoke them if needed. 4) If you prefer to avoid remote services, consider local HTML-to-image tools or a vetted package manager installation. If you want a deeper assessment, provide the actual installer script (https://cli.inference.sh) and the dist.inference.sh checksums page so I can inspect what's being downloaded and how verification is performed.
Capability Analysis
Type: OpenClaw Skill
Name: og-image-design
Version: 0.1.5
The skill is classified as suspicious primarily due to the `curl -fsSL https://cli.inference.sh | sh` command found in `SKILL.md`. This method of installing a CLI tool by piping a remote script directly to `sh` is a significant supply chain risk, as it allows for arbitrary code execution (RCE) on the agent's host if the `inference.sh` domain or its distribution server were to be compromised. While the skill's stated purpose of generating OG images is benign and other commands use the `infsh` tool as intended, this installation practice introduces a critical vulnerability.
Capability Assessment
Purpose & Capability
Name/description match the runtime instructions: the SKILL.md shows how to produce social/OG images by calling the infsh (inference.sh) HTML-to-image app. There are no unrelated env vars, binaries, or install requirements in the bundle.
Instruction Scope
The instructions explicitly tell the agent to download and run https://cli.inference.sh and to use `infsh app run ...` with inlined HTML. The guidance does not request reading local files or unrelated environment variables. However, the examples will transmit the provided HTML/content to inference.sh (a remote service) which has privacy implications; the SKILL.md does not detail what is sent, retained, or logged by that service.
Install Mechanism
There is no formal install spec in the skill bundle, but the SKILL.md advises piping a remote script into sh (curl | sh) that downloads a binary from dist.inference.sh. While the document claims checksum verification is available, executing a remote install script is higher risk than a vetted package manager; users should manually verify checksums and the publisher before running the installer.
Credentials
The skill declares no required env vars or credentials. In practice `infsh login` implies interactive authentication or tokens stored outside the skill, but the SKILL.md does not document required credentials or how they are stored. That omission is not necessarily malicious but users should expect to provide/authorize credentials for the remote service.
Persistence & Privilege
The skill does not request permanent presence (always:false), contains no code files, and does not modify other skills or system-wide settings. Autonomous invocation is allowed (default) but is appropriate for an instruction-only helper.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install og-image-design - After installation, invoke the skill by name or use
/og-image-design - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.1.5
- Expanded documentation with detailed platform specs, templates, and design best practices for OG/social sharing images.
- Added quick start guide demonstrating image generation via inference.sh CLI.
- Included HTML/CSS examples for blog, product, and tutorial OG images.
- Provided universal image size and layout recommendations for compatibility.
- Added OG meta tag and Twitter Card reference for proper SEO/social preview integration.
- Outlined safe zone, font, color, and layout rules for consistent, on-brand images.
v0.1.0
Initial release of og-image-design.
- Provides platform-specific templates and best practices for Open Graph/social sharing images.
- Includes universal dimensions, text placement, branding guidance, and safe zones.
- Offers HTML-to-image and AI-generated visual templates via inference.sh CLI.
- Documents OG/Twitter meta tags and platform requirements.
- Features design rules for text, color, and consistency across social previews.
Metadata
Frequently Asked Questions
What is Og Image Design?
Open Graph and social sharing image design with platform specs, text placement, and branding. Covers OG meta tags, Twitter cards, LinkedIn previews, and dyna... It is an AI Agent Skill for Claude Code / OpenClaw, with 815 downloads so far.
How do I install Og Image Design?
Run "/install og-image-design" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Og Image Design free?
Yes, Og Image Design is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Og Image Design support?
Og Image Design is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Og Image Design?
It is built and maintained by Ömer Karışman (@okaris); the current version is v0.1.5.
More Skills