← 返回 Skills 市场
Skill Scoreboard
作者
luis1213899
· GitHub ↗
· v1.3.0
· MIT-0
93
总下载
0
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install skill-scoreboard
功能描述
技能使用积分榜 v1.3 — OpenClaw 技能使用追踪与积分管理系统。 触发场景: (1) 用户询问技能使用榜单、积分统计(日榜/周榜/月榜/年榜/总榜) (2) 用户要求查看技能调用记录、错误日志 (3) 用户要求查看某技能的使用详情 (4) 用户要求生成每日/历史积分报告 (5) 用户询问"今天情况如何...
安全使用建议
This skill appears to implement the advertised leaderboard functionality, but note two important issues before installing: (1) An undocumented script (scripts/parse_gateway_logs.py) will read gateway logs from /tmp/openclaw and extract message snippets, then call the tracker — that may capture and persist sensitive user messages; (2) the tracker scans your SharedSkills directory to compute complexity, which can expose file contents and file counts. Actions to consider: (a) inspect parse_gateway_logs.py and score_tracker.py yourself (they are included) and confirm you are comfortable with them reading /tmp/openclaw and SharedSkills; (b) if you want only manual tracking, avoid running parse_gateway_logs.py and do not enable any automated cron; (c) run the scripts in a restricted/sandboxed environment or limit file permissions for /tmp/openclaw and SharedSkills if you must; (d) ask the publisher to update SKILL.md to disclose automatic log parsing and provide opt-in/opt-out controls. If you need help reviewing specific lines or tailoring the scripts to avoid reading sensitive paths, provide guidance and I can suggest edits.
功能分析
Type: OpenClaw Skill
Name: skill-scoreboard
Version: 1.3.0
The skill bundle implements a legitimate usage tracking and scoring system for the OpenClaw environment. It analyzes gateway logs in /tmp/openclaw to detect skill activity and calculates 'complexity weights' by scanning file metadata in the SharedSkills directory. All data is stored locally in ~/.skill_scoreboard, and no evidence of data exfiltration, unauthorized network access, or malicious command execution was found in scripts/score_tracker.py or scripts/parse_gateway_logs.py.
能力评估
Purpose & Capability
The stated purpose (track skill usage, compute scores, produce leaderboards) aligns with the included score_tracker.py which scans SharedSkills and writes ~/.skill_scoreboard. However the package also contains parse_gateway_logs.py (not mentioned in SKILL.md) which actively reads gateway logs from /tmp/openclaw and invokes score_tracker.py; this extra component is plausible for automatic tracking but wasn't documented in the runtime instructions.
Instruction Scope
SKILL.md documents only score_tracker usage and cron jobs; it does not mention parse_gateway_logs.py. parse_gateway_logs.py will read Gateway log files at /tmp/openclaw/openclaw-YYYY-MM-DD.log, extract message text (examples), and call score_tracker via subprocess. That means the skill will parse potentially sensitive user messages and logs without the SKILL.md disclosure. There are no network exfiltration calls in the code, but local log reading and example extraction are privacy-sensitive and broader than the documented commands.
Install Mechanism
This is an instruction-only skill with no install spec. No remote downloads or package installs are performed by the skill files themselves — the lowest-risk installation model.
Credentials
The skill requests no environment variables or credentials. It does read local filesystem locations: the SharedSkills directory (to compute complexity across all skills), and /tmp/openclaw logs (gateway logs). Reading SharedSkills and home-dir files is coherent with complexity-weighting and local storage, but it grants broad read access to user skill files and system logs which can contain secrets or sensitive content.
Persistence & Privilege
The skill does not request always:true and does not modify other skills. It creates ~/.skill_scoreboard and log files in the user's home directory (normal for a local tracker) and suggests adding cron entries (user-controlled). This is typical and proportional.
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 skill-scoreboard - 安装完成后,直接呼叫该 Skill 的名称或使用
/skill-scoreboard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
v1.3.0
新增五种榜单模式:日榜、周榜、月榜、年榜、总榜。修复今日榜成功率显示问题。
v1.2.0
v1.2: 修复文档不一致问题,移除不存在的Feishu推送依赖,修正权重示例值
v1.1.0
v1.1: 新增每日工作流复盘功能,自动找出积分冠军技能并固化最佳实践
元数据
常见问题
Skill Scoreboard 是什么?
技能使用积分榜 v1.3 — OpenClaw 技能使用追踪与积分管理系统。 触发场景: (1) 用户询问技能使用榜单、积分统计(日榜/周榜/月榜/年榜/总榜) (2) 用户要求查看技能调用记录、错误日志 (3) 用户要求查看某技能的使用详情 (4) 用户要求生成每日/历史积分报告 (5) 用户询问"今天情况如何... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 93 次。
如何安装 Skill Scoreboard?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install skill-scoreboard」即可一键安装,无需额外配置。
Skill Scoreboard 是免费的吗?
是的,Skill Scoreboard 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
Skill Scoreboard 支持哪些平台?
Skill Scoreboard 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 Skill Scoreboard?
由 luis1213899(@luis1213899)开发并维护,当前版本 v1.3.0。
推荐 Skills