interview-me for Openclaw

Prerequisites

• An AI agent platform that supports SKILL.md, such as Claude Code, Cursor, Codex, or Gemini CLI.
• Windows, macOS, or Linux.
• Access to an LLM API supplied by the agent platform.
• A required data source, described by the deployment or GitHub-based source configuration.
• Prepared input content in JSON, text, or Markdown format.

Install and register the skill

The source document does not specify a package-manager installation command. Install or copy the interview-me skill directory into the skills location supported by your agent, ensuring that SKILL.md remains at the expected path.

mkdir -p ~/.openclaw/skills/interview-me
cp SKILL.md ~/.openclaw/skills/interview-me/SKILL.md

Register or reload the skill from your agent's configuration, then verify that the agent can discover interview-me and its read and exec tool permissions.

Configure credentials

Provide the LLM API credentials through the agent platform's supported secret or environment-variable mechanism. Do not commit API keys to the skill directory.

export LLM_API_KEY="your-api-key"

Invoke the skill with required content and optional format and options values. Confirm that the source data is reachable, permissions are correct, and logs are enabled for troubleshooting and auditing.

Request parameters

Response envelope

{
  "success": true,
  "data": {
    "result": "Clarified and structured interview-me result",
    "metadata": {
      "skill": "interview-me",
      "version": "1.0.0",
      "pricing_tier": "L2-advanced"
    }
  },
  "error": null
}

Data organization

• content is the source request being analyzed.
• format identifies how the input should be parsed and normalized.
• options stores optional execution preferences and batch-processing controls.
• data.result contains the clarified, structured output produced by the interview workflow.
• data.metadata records skill identity, version, and pricing tier for traceability.
• error is null for successful execution and contains failure information when processing does not complete.
• No specific output filenames or persistent storage layout are defined in the source document; file paths should therefore be supplied by the host agent or integration.