Model Benchmark

深度测评各模型在 OpenClaw 上的实际表现，支持中文理解/代码/推理/工具调用多维度评估。

This skill appears to be a legitimate benchmarking instruction set, but it refers to obtaining and using multiple external provider API keys without declaring them in the metadata or describing how to provide or store them. Before installing or using it: (1) Confirm how you'll supply provider keys — prefer ephemeral or least-privilege keys and avoid pasting long-lived secrets into third-party UIs. (2) Understand where keys will be stored (models.json) and check file permissions; back up the original config. (3) Verify the local proxy address (127.0.0.1:8766) is expected in your environment and not a misdirection to an unfamiliar service. (4) If the agent will be given keys, ensure the agent's prompts/storage behavior is acceptable (won't exfiltrate them). If you need complete assurance, request the skill author to declare required env vars and document exactly how credentials are used and persisted.

Type: OpenClaw Skill Name: model-benchmark Version: 0.1.0 The skill bundle is a benchmarking framework designed to evaluate AI models on the OpenClaw platform across dimensions like code generation, reasoning, and tool use. The files (SKILL.md and skill.md) contain only descriptive instructions, test cases, and reporting templates without any executable code, data exfiltration logic, or malicious prompt injections. It correctly directs the user to configure API keys through the platform's standard 'models.json' configuration.