Model Config Check

校验模型配置是否正确、模型是否可以正常连接和返回内容。当用户说"检查模型"、"测试模型"、"模型能不能用"、"模型配置"、"诊断模型问题"时使用。**每次修改模型配置（config.patch/config.apply涉及models.providers）后必须自动执行校验。** 用户只给模型名+API key时...

This skill appears to do what it says (check model provider configs and call providers to verify responses) but has a few red flags you should address before installing: (1) The SKILL.md promises auto-detect/auto-write of provider config when a user gives only model+API key, but the included script does not implement writing/applying configs — clarify or remove the promise. (2) The script uses python3 and curl at runtime but the skill metadata lists no required binaries; ensure those tools exist on hosts where it will run. (3) The extractor prints API keys into its tab-separated output, which can leak keys to stdout/logs — restrict log access or modify the script to avoid printing secrets. (4) The skill expects to be auto-run on gateway config changes; if you enable automatic invocation, ensure it runs under appropriate permissions and that audit/logging is in place. Recommended next steps: review and harden the script to avoid printing secrets, add declared required binaries, test in a staging environment, and only enable automatic triggers after you confirm safe invocation hooks and logging controls.

Type: OpenClaw Skill Name: model-config-check Version: 1.0.0 The skill bundle is designed to validate AI model configurations and connectivity. It provides instructions for the agent to automatically verify settings after configuration changes and includes a shell script (check_models.sh) that uses curl and Python to test API endpoints. While the skill handles sensitive data like API keys and performs network requests, these actions are strictly aligned with its stated purpose of diagnosing model issues, and there is no evidence of data exfiltration to unauthorized third parties or malicious execution logic.