pasteclaw-agent

What This Skill Does

The Pasteclaw Agent Skill provides a seamless interface for interacting with Pasteclaw.com, enabling AI agents to publish, update, and organize text-based content directly from their execution environment. It acts as an "agent-friendly" clipboard, supporting various formats including HTML, CSS, JavaScript, Markdown, JSON, and YAML. By providing preview URLs for generated assets, it allows agents to share visual prototypes, documentation, and data snapshots instantly with users or other agents. The skill includes support for custom meta headers, allowing for detailed audit trails of which AI model or tool generated the content.

Installation

You can integrate this skill into your OpenClaw environment using the following command: clawhub install openclaw/skills/skills/tairov/pasteclaw Ensure your agent environment has network access to https://pasteclaw.com to successfully transmit payloads.

Use Cases

• Prototyping: Generate functional HTML/CSS/JS components and share live previews immediately.
• Data Exchange: Export complex JSON or YAML outputs as shareable URLs for review by human operators or other agents.
• Documentation: Automatically generate and host Markdown documentation or technical READMEs on the fly.
• Agent Collaboration: Use Pasteclaw as a shared workspace/buffer to pass structured data between multiple autonomous agents.
• Feedback Loops: Enhance user trust by sharing "show your work" artifacts that demonstrate the results of complex tasks.

Example Prompts

1. "Create a Pasteclaw snippet titled 'Dashboard UI' using the HTML and CSS I just generated, and return the shareable URL."
2. "Save this JSON configuration to Pasteclaw under a new snippet, tag it as 'tool=data-analyzer', and provide me the public link."
3. "Summarize the project findings in Markdown format and publish them to Pasteclaw, ensuring the meta-header marks the task as 'source=research-agent'."

Tips & Limitations

• Traceability: Always include the X-Pasteclaw-Meta header to identify which model or tool created the content. This is invaluable for long-term project auditing.
• Content Types: Explicitly define the contentType (e.g., 'markdown', 'json') when posting raw text to ensure the preview renders correctly.
• Persistence: Pasteclaw is intended for sharing and previewing; while it provides stable URLs, ensure sensitive data is not shared if privacy is a concern.
• Fallback: If your execution environment lacks curl, utilize the provided Python snippet to achieve the same result without needing external binary dependencies.