Skill Tracker

v1.2.0

通用技能使用统计追踪器,支持 Python 和 Node.js 技能,自动记录调用次数、成功率,生成使用排行榜。数据本地存储,保护隐私。

0· 168·1 current·1 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for aiyst1982/skill-tracker-aiyst.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Skill Tracker" (aiyst1982/skill-tracker-aiyst) from ClawHub.
Skill page: https://clawhub.ai/aiyst1982/skill-tracker-aiyst
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Required binaries: python3
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install skill-tracker-aiyst

ClawHub CLI

Package manager switcher

npx clawhub@latest install skill-tracker-aiyst
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
medium confidence
Purpose & Capability
The name/description match the code: both Python and Node.js implementations are present and implement local usage tracking. Minor inconsistencies: registry metadata and SKILL.md declare python3 as the required binary but a full Node.js implementation (index.js, package.json) is included — if you plan to use the Node version you must ensure Node is available. The registry also labeled this as 'instruction-only' despite shipping code files; this is plausibly just packaging but worth noting.
Instruction Scope
SKILL.md and the code only write local files (skill-stats.json, usage-log.jsonl) and provide APIs for integrators to call. They record fields like user, channel, session_id and error text which can contain PII or session identifiers. The docs reference a script (node scripts/check-integration.js) that is not present in the package. There are no instructions to read unrelated system files or send data externally.
Install Mechanism
No remote download/install mechanism is declared; repository includes source files (Python/Node) and setup.py / package.json for local install. No suspicious URLs, extract steps, or third-party binaries observed. This is low-risk from an install-source perspective.
Credentials
The skill requests no environment variables or credentials (proportionate). However, it logs context fields (user, session_id, channel, error) that can contain sensitive data; configuration allows changing data_dir (examples show /var/log/skill-tracker), which could require elevated privileges if set to system locations. No external credentials are requested.
Persistence & Privilege
The skill persists data locally and does not request elevated platform privileges. always is false and it does not modify other skills' configs. Be aware that changing the configured data_dir to system paths (e.g., /var/log) may require higher privileges and could increase persistence/impact.
Assessment
This package appears to implement a local-only skill analytics tracker, but before installing: 1) Verify the source (GitHub homepage) and that you trust the publisher. 2) Confirm which implementation you will use (Python vs Node); if you use the Node version, ensure Node is available (metadata only declared python3). 3) Inspect/choose the data_dir: default is package-relative data/ (safe), but examples use /var/log which may require root—don't set system directories unless intended. 4) Review what you log: the tracker saves user, session_id, channel and error text — these can be PII or session tokens. Turn off logRaw or redact sensitive fields if needed and set an appropriate retention policy. 5) Note a small docs inconsistency (referenced script missing); if you rely on any helper scripts, verify they exist. 6) If you need stronger guarantees, run the code in a sandbox or review the code paths that write files to ensure no unexpected file access or network calls are introduced in future versions.

Like a lobster shell, security has layers — review code before you run it.

Runtime requirements

📊 Clawdis
OSLinux · macOS · Windows
Binspython3
latestvk97ddfjn6w35cz7c926vvk8ts183741g
168downloads
0stars
1versions
Updated 1mo ago
v1.2.0
MIT-0
Linux, macOS, Windows

Skill Tracker 📊

通用技能使用统计追踪器 - 让您的 OpenClaw 技能使用情况一目了然!


何时使用

当您需要:

  1. 知道哪个技能最常用
  2. 发现哪个技能总失败
  3. 识别哪些技能可以吃灰
  4. 数据驱动优化技能配置
  5. 监控技能使用趋势

典型场景:

  • "我最近都用哪些技能?"
  • "哪个技能调用次数最多?"
  • "有没有技能总失败需要检查?"
  • "哪些技能可以卸载了?"

触发场景

自动触发

  • 每次技能调用时自动记录(需技能已集成)
  • 无需手动指令

手动查询

指令:技能统计
指令:调用统计
指令:skill 排行
指令:使用报告
指令:哪个技能最常用
指令:查看技能使用情况

功能特性

核心功能

  • 自动记录 - 调用/成功/失败自动统计
  • 使用排行 - 按调用次数排序
  • 成功率统计 - 跟踪每个技能稳定性
  • 时间分析 - 按日期统计
  • 双语言支持 - Python + Node.js
  • 本地存储 - 数据安全
  • 零外部依赖 - 纯本地运行

统计维度

  • 总调用次数
  • 技能总数
  • 每个技能的调用/成功/失败次数
  • 成功率百分比
  • 首次/最后使用时间

使用方法

查询统计

指令:技能统计

返回示例:

📊 技能使用统计报告
==================================================
总调用次数:156
技能总数:11

技能排行榜(按调用次数):

🥇 **1. nano-banana-pro**
   调用:45 | 成功:44 | 失败:1 | 成功率:97.8%

🥈 **2. bailian-usage**
   调用:32 | 成功:32 | 失败:0 | 成功率:100%

🥉 **3. fact-checker**
   调用:28 | 成功:28 | 失败:0 | 成功率:100%

集成到您的技能

Python 技能:

import sys
sys.path.insert(0, '../skill-tracker')
from skill_tracker import track

# 记录调用
track('your-skill', 'call')
track('your-skill', 'success')
track('your-skill', 'fail', {'error': 'xxx'})

Node.js 技能:

const tracker = require('../skill-tracker');

await tracker.track('your-skill', 'call', { context });
await tracker.track('your-skill', 'success', { context });
await tracker.track('your-skill', 'fail', { context, error });

配置

可选配置文件:skills/skill-tracker/config.yaml

tracker:
  enabled: true          # 是否启用追踪
  logRaw: true          # 是否记录原始日志
  retentionDays: 90     # 日志保留天数
  
report:
  topN: 10             # 默认显示前几名
  includeDetails: true # 包含详细信息

数据文件

文件位置说明
skill-stats.jsonskills/skill-tracker/data/聚合统计
usage-log.jsonlskills/skill-tracker/data/原始日志

隐私说明:

  • ✅ 所有数据本地存储
  • ✅ 不上传到外部服务器
  • ✅ 可随时删除日志文件

已集成的技能(示例)

技能类型用途
nano-banana-proPythonAI 图片生成
taobaoPython淘宝商品搜索
libtv-skillPythonAI 视频生成
baidu-searchPython百度搜索
feastPython膳食规划
evernote-skillNode.js印象笔记
fact-checkerNode.js观点追踪
bailian-usageNode.js阿里云用量查询

命令行使用

# 查看统计报告
cd skills/skill-tracker
python3 skill_tracker.py report

# 测试追踪器
python3 skill_tracker.py test

# 检查集成状态
node scripts/check-integration.js

常见问题

Q: 会影响技能性能吗?
A: 几乎无影响,每次调用增加 <1ms 延迟。

Q: 数据会上传吗?
A: 不会,所有数据存储在本地。

Q: 如何清理数据?
A: 删除 skills/skill-tracker/data/ 目录即可。

Q: 支持自定义技能吗?
A: 支持!按文档添加几行代码即可。


技术规格

  • 支持语言: Python 3.6+, Node.js 14+
  • 支持系统: Linux, macOS, Windows
  • 外部依赖:
  • 网络访问: 不需要
  • 权限要求: 文件系统读写

更新日志

v1.0.0 (2026-03-18)

  • ✨ 初始版本发布
  • ✨ Python + Node.js 双版本支持
  • ✨ 已集成 11 个技能
  • ✨ 本地存储,零外部依赖
  • ✨ 命令行 + 对话两种查询方式

许可证

MIT License


作者

韩顺利 (@hanshunli)


📊 让数据驱动您的技能优化决策!

Comments

Loading comments...