← Back to Skills Marketplace
dingtom336-gif

Skill Scorer

by dingtom336-gif · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
141
Downloads
0
Stars
0
Active Installs
3
Versions
Install in OpenClaw
/install skill-scorer
Description
对任何 SKILL.md(或 skill 文件夹)进行质量评估和打分,基于行业最佳实践,生成 8 维度 100 分制的结构化质检报告,精准定位问题并提供可执行的优化建议。当用户要求评审、审计、评分、检测、质检任何 skill 时使用——哪怕只是说「这个 skill 写得怎么样?」也会触发。也支持:skill质检、...
Usage Guidance
This skill appears coherent and safe for its stated purpose: auditing SKILL.md files. Things to consider before installing or running it: (1) it will fetch and evaluate user-provided URLs/GitHub links — only provide links you trust and be aware the agent may retrieve remote content; (2) the skill can generate 'auto-fix' versions of SKILL.md — always review any generated changes before applying them; (3) although it requests no secrets, avoid pasting sensitive credentials or private keys into files you ask it to fetch or evaluate. If you want higher assurance, ask the maintainer to explicitly document how fetched remote content is handled (read-only analysis vs. execution) and to add a reminder not to include secrets in submitted skill artifacts.
Capability Analysis
Type: OpenClaw Skill Name: skill-scorer Version: 1.0.0 The skill-scorer bundle is a meta-skill designed to audit and score the quality of other OpenClaw skills. It functions as a linter and quality-assurance tool, evaluating SKILL.md files across eight dimensions such as metadata, workflow logic, and safety. The logic contained in SKILL.md and its reference files (rubric.md, anti-patterns.md) is transparent, well-documented, and lacks any indicators of data exfiltration, malicious execution, or prompt injection. It even includes specific instructions to validate input and avoid processing non-skill files, demonstrating a focus on robustness and intended use.
Capability Assessment
Purpose & Capability
Name, description, and bundled reference files (rubric, anti-patterns, report template) match the advertised purpose of auditing SKILL.md files; required resources (none) are proportionate to the task.
Instruction Scope
Runtime instructions are focused on parsing and scoring skill files and references. One minor scope note: Step 0 says 'URL / GitHub link → Fetch and evaluate', which implies making outbound requests to user-provided URLs; this is reasonable for a reviewer but the skill does not enumerate any constraints or verification steps for remote content (e.g., do not execute fetched scripts). Also the report-template includes an auto-fix flow ('fix' / 'rewrite') that will generate new SKILL.md content — expected for this meta-skill but users should review generated edits before applying.
Install Mechanism
Instruction-only skill with no install spec and no bundled executables; nothing is written or downloaded by an installer. README shows manual copy instructions (user-facing), which is normal.
Credentials
No environment variables, credentials, or config paths are requested. The skill does mention reading files in a provided skill folder and fetching user-provided URLs — appropriate for this purpose and not disproportionate.
Persistence & Privilege
always:false and user-invocable defaults; the skill does not request persistent privileges or to modify other skills or system-wide settings. Its described auto-fix behaviors generate content but do not claim to self-enable or persist beyond normal agent invocation.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install skill-scorer
  3. After installation, invoke the skill by name or use /skill-scorer
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
1.0.0 — Initial release with an 8-dimension scoring rubric: - Evaluates any SKILL.md (or skill folder) for quality and compliance. - Assigns scores out of 100 across 8 weighted dimensions based on best practices. - Identifies all issues by severity with actionable, specific suggestions. - Generates a bilingual (Chinese + English) structured quality report for every evaluation.
v1.6.0
v1.6.0
v1.5.0
v1.5.0 release
Metadata
Slug skill-scorer
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 3
Frequently Asked Questions

What is Skill Scorer?

对任何 SKILL.md(或 skill 文件夹)进行质量评估和打分,基于行业最佳实践,生成 8 维度 100 分制的结构化质检报告,精准定位问题并提供可执行的优化建议。当用户要求评审、审计、评分、检测、质检任何 skill 时使用——哪怕只是说「这个 skill 写得怎么样?」也会触发。也支持:skill质检、... It is an AI Agent Skill for Claude Code / OpenClaw, with 141 downloads so far.

How do I install Skill Scorer?

Run "/install skill-scorer" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Skill Scorer free?

Yes, Skill Scorer is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Skill Scorer support?

Skill Scorer is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Skill Scorer?

It is built and maintained by dingtom336-gif (@dingtom336-gif); the current version is v1.0.0.

💬 Comments