← 返回 Skills 市场
Performance Tuning
作者
mikeclaw007
· GitHub ↗
· v1.0.0
· MIT-0
251
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install performance-tuning
功能描述
Deep performance tuning workflow—goals and measurement, profiling, hotspots, caching and concurrency trade-offs, system-specific tuning (DB, GC, network), an...
安全使用建议
This skill is a benign, high-level checklist for performance tuning and does not request credentials or install anything. Consider these practical points before enabling it: (1) it provides advice only — it will not run profilers or change systems unless paired with other skills or human-invoked actions; (2) because the source is unknown, avoid pasting sensitive production traces or secrets into prompts when seeking guidance; (3) if you later augment the skill with automation (scripts, installs, or profiling commands), re-evaluate for permissions and network/download risks; (4) test recommendations in a safe environment (staging/canary) before applying to production.
功能分析
Type: OpenClaw Skill
Name: performance-tuning
Version: 1.0.0
The skill bundle contains a high-level procedural workflow for performance tuning and optimization. It consists entirely of documentation and instructions for an AI agent to follow (SKILL.md) and standard metadata (_meta.json), with no executable code, suspicious network calls, or attempts to access sensitive system resources.
能力评估
Purpose & Capability
Name and description match the provided content. The skill is an advisory workflow and requests no credentials, binaries, or config access — which is appropriate for a guidance-only performance-tuning skill.
Instruction Scope
SKILL.md contains step-by-step, measurement-driven guidance and tool suggestions but does not instruct the agent to read local files, access environment variables, call external endpoints, download or execute code, or collect/transmit sensitive data. It stays within the stated advisory scope.
Install Mechanism
No install spec and no code files — instruction-only. This minimizes disk writes and runtime installation risk.
Credentials
No environment variables, credentials, or config paths are required or referenced. The lack of requested secrets is proportionate to an advisory workflow.
Persistence & Privilege
always is false and the skill does not request persistent system presence or modifications to other skills. Autonomous invocation is allowed by default but the skill's content is advisory and does not perform privileged operations.
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 performance-tuning - 安装完成后，直接呼叫该 Skill 的名称或使用
/performance-tuning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
v1.0.0
Initial release of the performance-tuning skill.
- Introduces a six-stage, measurement-driven workflow for deep performance tuning.
- Covers goal setting, baseline measurement, profiling, implementation, verification, and regression prevention.
- Includes supporting checklists, profiling tool examples, and tips for various system types (e.g., DB, GC, network).
- Designed for addressing latency, throughput, resource saturation, or efficiency/cost concerns.
元数据
常见问题
Performance Tuning 是什么？
Deep performance tuning workflow—goals and measurement, profiling, hotspots, caching and concurrency trade-offs, system-specific tuning (DB, GC, network), an... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 251 次。
如何安装 Performance Tuning？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install performance-tuning」即可一键安装，无需额外配置。
Performance Tuning 是免费的吗？
是的，Performance Tuning 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Performance Tuning 支持哪些平台？
Performance Tuning 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 Performance Tuning？
由 mikeclaw007（@mikeclaw007）开发并维护，当前版本 v1.0.0。
推荐 Skills