AgentVibes Content Creation and Management Skill for Openclaw

Prerequisites

• A compatible SKILL.md-based AI agent, including Claude Code, Cursor, Codex, or Gemini CLI.
• Windows, macOS, or Linux.
• Access to an LLM provided by the host agent.
• An AgentVibes-compatible API key and network access to the required external service.
• Input data prepared in a supported format. The documented language support is English and Simplified Chinese.

Install or enable the skill

Place the skill definition in the skills directory recognized by your AI agent, or enable it through the agent platform's skill-management workflow. The source document does not define a standalone package manager command or a separate dependency installation script.

Configure credentials

Set the API key as an environment variable rather than hard-coding it in prompts, source files, or version control:

export API_KEY="${API_KEY:?Please set the API_KEY environment variable}"

Restart the current agent session or open a new terminal after configuration so the environment is available. Keep the key secret, use HTTPS for API traffic, and restrict access to authorized users.

First request

Provide a required instruction and, when useful, additional context in the agent conversation. A typical logical request contains instruction as a string and optional context as a string. Confirm that the API key, network connection, input size, and user permissions are valid before starting a production workflow.

Request schema

Response schema

{
  "success": true,
  "data": {
    "result": "Generated or processed skill result"
  },
  "error": null
}

The documented example contains a duplicated result key; integrations should treat data.result as the canonical output field and validate the response before downstream processing.

Content and metadata organization

• Content task: instruction, optional context, requested format, and target publication channel.
• Media scope: text, image, audio, HTML, Markdown, or PDF-oriented output, subject to the connected AgentVibes API.
• Operational metadata: success state, result payload, error details, latency, retry status, and rate-limit events.
• Governance metadata: tenant identity, user role, permission decisions, audit records, and compliance logs.
• Configuration metadata: API credentials, retry policy, timeout values, platform settings, and agent environment.
• Files: the skill can read, process, and write supported files using the host agent's read and write tools; the source does not prescribe a fixed output directory or filename convention.

Documented limits

• Maximum input content per API request: 10,000 characters in the general API boundary specification.
• Uploaded image, video, or audio file size: up to 50 MB.
• Maximum concurrent requests: 100.
• API rate limit: up to 20 calls per second.
• Stated free API-key allowance: 10,000 calls per month.
• Tenant creation and permission changes require administrator access.
• The documented AgentVibes feature version requirement is 4.6.7 or later.

The document also lists a separate optimization profile with an average response target below 2 seconds, at least 10 concurrent requests, memory usage below 100 MB, input below 10 MB, and an error rate below 1%. Treat these as stated performance targets rather than universal guarantees, because actual results depend on the connected model, API service, workload, and network conditions.