← 返回 Skills 市场
MetriLLM
作者
TheBlueHouse75
· GitHub ↗
· v0.2.11
427
总下载
0
收藏
0
当前安装
11
版本数
在 OpenClaw 中安装
/install metrillm
功能描述
Find the best local LLM for your machine. Tests speed, quality and RAM fit, then tells you if a model is worth running on your hardware.
安全使用建议
This skill is coherent for benchmarking local LLMs, but take two precautions before installing: (1) review the npm package / GitHub repo (https://github.com/MetriLLM/metrillm) or audit the package contents before running npm install -g, since global npm installs execute third-party code on your machine; (2) only use --share if you consent to publishing model names, scores and hardware details (the README says no personal data is sent, but verify what the package actually uploads). Also ensure you have Node 20+ and run Ollama or LM Studio locally as instructed.
功能分析
Type: OpenClaw Skill
Name: metrillm
Version: 0.2.11
The skill bundle contains a shell injection vulnerability in SKILL.md by instructing the agent to pass unvalidated $ARGUMENTS directly into Bash commands (e.g., `metrillm bench --model $ARGUMENTS`). It also performs high-risk operations including a global npm installation and optional data exfiltration of hardware and performance metrics to an external endpoint (metrillm.dev). While these actions are documented and align with the tool's purpose as a benchmark utility, the combination of broad Bash permissions and the injection risk poses a significant security threat.
能力评估
Purpose & Capability
The name/description match the instructions: it tells you how to install the metrillm CLI, requires Node 20+ and a local LLM server (Ollama or LM Studio), and runs benchmarking commands. No unrelated credentials, binaries, or config paths are requested.
Instruction Scope
Instructions stay within the benchmarking scope (run metrillm bench, view local ~/.metrillm/results/). One caution: the optional --share command uploads results (model name, scores, hardware specs) to metrillm.dev; the SKILL.md states no personal data is sent, but that claim cannot be verified from instructions alone. The skill does not instruct access to unrelated files or env vars.
Install Mechanism
Installation is via npm (npm install -g metrillm) which is a standard delivery for a Node CLI. npm installs are moderate-risk because they execute third-party code on your system; this is proportionate to the stated purpose but you should inspect the package or source repository before global installation.
Credentials
No environment variables, credentials, or config paths are required. The only data potentially exported is from the explicit --share action (model, scores, hardware specs), which is reasonable for a community leaderboard.
Persistence & Privilege
The skill is not always-enabled and is user-invocable. It does not request persistent elevated privileges or modify other skills. Autonomous invocation is permitted by default (normal), but nothing in the skill attempts to gain extra persistence.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install metrillm - 安装完成后，直接呼叫该 Skill 的名称或使用
/metrillm触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v0.2.11
Fix license: Apache-2.0, not MIT
v0.2.10
Add open source section with GitHub repo link
v0.2.9
Use global metrillm commands instead of npx, remove duration estimates
v0.2.8
Use global metrillm commands instead of npx, remove duration estimates
v0.2.7
Remove misleading benchmark duration estimates
v0.2.6
Fix install command: npm install -g instead of npx
v0.2.5
Add install instructions, LM Studio support, community leaderboard description, --share transparency for security scan compliance
v0.2.3
Rewrite description for clarity
v0.2.2
Improve listing description
v0.2.1
Security fixes, zero-config MCP, LM Studio improvements
v0.1.1
- Initial public release of metrillm.
- Benchmark local LLM models for both performance (tokens/sec, time to first token, memory) and quality (reasoning, math, coding, etc.).
- Computes clear hardware fitness verdict: EXCELLENT / GOOD / MARGINAL / NOT RECOMMENDED.
- Works with models run locally through Ollama.
- Supports quick performance-only benchmarks and sharing results to a public leaderboard.
元数据
常见问题
MetriLLM 是什么？
Find the best local LLM for your machine. Tests speed, quality and RAM fit, then tells you if a model is worth running on your hardware. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 427 次。
如何安装 MetriLLM？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install metrillm」即可一键安装，无需额外配置。
MetriLLM 是免费的吗？
是的，MetriLLM 完全免费（开源免费），可自由下载、安装和使用。
MetriLLM 支持哪些平台？
MetriLLM 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 MetriLLM？
由 TheBlueHouse75（@thebluehouse75）开发并维护，当前版本 v0.2.11。
推荐 Skills