Token导航 LogoToken导航TokenDH.com
研究检索只读github未标认证来源可访问clear审计通过

system-reviewer系统审核员

Agent Skill

system-reviewer 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

356

周安装

15

GitHub Stars

9

下载量

125
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

3

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:system-reviewer(系统审核员)
来源仓库:https://github.com/adaptationio/skrillz
仓库路径:skills/system-reviewer
安装命令:
npx skills add https://github.com/adaptationio/skrillz --skill system-reviewer
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/adaptationio/skrillz --skill system-reviewer

简介

用于查找、筛选与系统审核相关的信息和标准。system-reviewer 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

  • 适合根据关键词快速定位合规要求、安全策略或质量准则。
  • 可辅助生成检查清单、识别潜在风险点或提供改进建议。
  • 使用时应结合具体领域规范,避免通用结论直接套用。
  • 建议人工复核关键判断,确保符合最新法规与技术标准。

SKILL.md

System Reviewer

Overview

system-reviewer performs meta-level reviews of the skill development ecosystem itself, assessing the health and effectiveness of the entire system rather than individual skills.

Purpose: Meta-level ecosystem assessment and system optimization

The 5 System Review Operations:

  1. Ecosystem Health Check - Assess overall ecosystem quality and completeness
  2. Toolkit Effectiveness Review - Evaluate how well the toolkit works
  3. Process Efficiency Review - Review development processes for optimization
  4. Coverage Gap Analysis - Identify missing capabilities or skills
  5. System Optimization Recommendations - Recommend system-level improvements

Key Distinction: Reviews the SYSTEM, not individual skills

When to Use

  • Periodic ecosystem health checks (monthly/quarterly)
  • After completing major layers (Layer 2, 3, 4, 5)
  • When suspecting systemic issues (not individual skill problems)
  • Planning next ecosystem developments
  • Optimizing the development system itself

Operations

Operation 1: Ecosystem Health Check

Purpose: Assess overall ecosystem quality, completeness, and consistency

Process:

  1. Count and Categorize Skills

- Total skills built - Skills per layer - Pattern distribution (workflow/task/reference) - Completion percentage

  1. Assess Structural Quality

- Run review-multi validation on all skills - Calculate average quality scores - Identify outliers (very high/low quality) - Check consistency across ecosystem

  1. Evaluate Completeness

- Are all planned layers complete? - Are there gaps in capabilities? - Is toolkit comprehensive? - Missing critical functionality?

  1. Check Integration

- Do skills work together well? - Are there integration issues? - Workflow compositions effective? - Dependencies clear and working?

  1. Generate Health Report

- Overall health score (healthy/good/needs attention/critical) - Strengths (what's working well) - Weaknesses (what needs improvement) - Trends (improving/stable/degrading)

Outputs:

  • Ecosystem health score
  • Skill inventory and metrics
  • Quality assessment aggregate
  • Completeness analysis
  • Integration assessment
  • Health report with recommendations

Time Estimate: 1-2 hours

Example:

Ecosystem Health Check: 2025-11-07
====================================

Skill Inventory:
- Total Skills: 23
- Layer 2: 10/10 (100%)
- Layer 3: 7/7 (100%)
- Layer 4: 6/6 (100%)
- Layer 5: 0/5 (0%)

Quality Assessment:
- Average Structure Score: 5.0/5.0 (all 23 skills Grade A)
- Quality Range: 5/5 (no variation - excellent consistency)
- Quick Reference Coverage: 100% (all 23 skills)
- Anti-Patterns: 10 total identified, 3 fixed

Completeness:
✅ Research capability: Complete (skill-researcher + research-workflow)
✅ Planning capability: Complete (planning-architect + task-development + planning-workflow)
✅ Execution capability: Complete (todo-management + momentum-keeper)
✅ Quality capability: Complete (review-multi + skill-validator + skill-tester + review-workflow)
✅ Improvement capability: Complete (skill-reviewer + skill-updater + improvement-workflow)
⬜ Self-improvement: 0% (Layer 5 not built)
⬜ Comprehensive testing: Partial (skill-tester exists, testing-validator missing)

Integration Assessment:
✅ Skills compose well (development-workflow, review-workflow, improvement-workflow working)
✅ Dependencies clear (YAML + documentation)
✅ Workflows effective (proven through usage)

Ecosystem Health: ✅ HEALTHY

Strengths:
- 100% structural excellence
- Complete Layers 2-4
- Validated continuous improvement cycle
- 85% efficiency gain proven

Weaknesses:
- Layer 5 incomplete (missing self-improvement automation)
- testing-validator missing (validation incomplete without it)
- Some minor anti-patterns in 4 skills (vague validation criteria)

Trends:
✅ Improving: Efficiency compounding, quality consistent, standards evolving
✅ Stable: Structural quality (all 5/5)

Recommendations:
1. [High] Complete Layer 5 for full self-improvement capability
2. [Medium] Build testing-validator for complete validation suite
3. [Low] Refine vague validation criteria in 4 skills

Operation 2: Toolkit Effectiveness Review

Purpose: Evaluate how well the development toolkit enables skill building

Process:

  1. Measure Efficiency Gains

- Build time progression (skills 1-23) - Efficiency vs baseline - Time saved calculation - Trend analysis (improving/plateauing?)

  1. Assess Tool Utilization

- Which tools used most? (high value) - Which tools rarely used? (low value or unknown) - Which tools most effective? (biggest impact) - Are there gaps? (missing tools)

  1. Evaluate Workflow Effectiveness

- Is development-workflow actually used? - Does it save time as promised? - Are workflows followed or bypassed? - Improvements to workflows needed?

  1. Check Tool Quality

- Are tools themselves high quality? - Do they achieve stated purposes? - User satisfaction with tools? - Tools needing improvement?

  1. Generate Toolkit Assessment

- Most valuable tools - Least utilized tools - Effectiveness metrics - Improvement opportunities

Outputs:

  • Toolkit effectiveness assessment
  • Tool utilization metrics
  • High-value vs low-value tools identified
  • Toolkit improvement recommendations

Time Estimate: 1-1.5 hours


Operation 3: Process Efficiency Review

Purpose: Review development processes for optimization opportunities

Process:

  1. Document Current Processes

- How are skills currently built? - What's the typical workflow? - What steps are always followed? - What steps are sometimes skipped?

  1. Measure Process Metrics

- Average build time per skill complexity - Rework rate (how often redo work?) - Blocker frequency (how often stuck?) - Completion rate (how many finished?)

  1. Identify Bottlenecks

- Longest steps in process - Most frequent blockers - Highest rework areas - Inefficient patterns

  1. Assess Automation

- What's automated? (scripts, workflows) - What's manual? (still required human input) - Automation opportunities? (could be automated) - Automation effectiveness? (actually saves time?)

  1. Generate Process Recommendations

- Process simplifications - Automation opportunities - Bottleneck elimination - Efficiency improvements

Outputs:

  • Process efficiency assessment
  • Bottleneck analysis
  • Automation opportunities
  • Process optimization recommendations

Time Estimate: 1-2 hours


Operation 4: Coverage Gap Analysis

Purpose: Identify missing capabilities, skills, or functionality in ecosystem

Process:

  1. Review Original Plan

- What was planned? (39 skills originally) - What's built? (23 skills currently) - What's missing? (16 skills remaining)

  1. Assess Current Capabilities

- Research: Complete ✅ - Planning: Complete ✅ - Execution: Complete ✅ - Quality: Mostly complete (missing testing-validator) - Improvement: Complete ✅ - Automation: Partial (Layer 5 missing)

  1. Identify Critical Gaps

- Must-have missing capabilities - High-value skills not built - Ecosystem limitations without them

  1. Prioritize Remaining Work

- Critical (must build) - High value (should build) - Nice to have (optional) - Low priority (defer)

  1. Generate Coverage Report

- Gap analysis (what's missing) - Impact assessment (what's the cost of gap) - Prioritized build recommendations

Outputs:

  • Coverage gap analysis
  • Missing capabilities identified
  • Prioritized skill build recommendations

Time Estimate: 45-90 minutes


Operation 5: System Optimization Recommendations

Purpose: Recommend system-level improvements (not individual skill improvements)

Process:

  1. Aggregate Findings

- From Operations 1-4 - From analysis of all skills - From usage patterns - From feedback

  1. Identify System-Level Issues

- Structural issues (ecosystem organization) - Process issues (development workflow problems) - Tool issues (toolkit gaps or inefficiencies) - Integration issues (skills not composing well)

  1. Generate Recommendations

- System reorganization needs - Process improvements - Toolkit enhancements - Integration optimizations

  1. Prioritize by Impact

- Critical (affects ecosystem viability) - High (significantly improves system) - Medium (moderate improvement) - Low (nice to have)

  1. Create Action Plan

- What to build next - What to improve - What to refactor - What to deprecate (if anything)

Outputs:

  • System optimization recommendations
  • Prioritized action plan
  • Impact assessments
  • Next development priorities

Time Estimate: 1-2 hours


Best Practices

1. Regular Health Checks

Practice: Run ecosystem health check monthly or after major milestones

Rationale: Early detection of systemic issues prevents compounding

2. Data-Driven Assessment

Practice: Use metrics and evidence, not opinion

Rationale: Objective assessment enables better decisions

3. System-Level Focus

Practice: Focus on ecosystem patterns, not individual skill issues

Rationale: system-reviewer is for systemic issues, review-multi for individual skills

4. Act on Findings

Practice: Use recommendations to guide next development priorities

Rationale: Reviews without action don't improve anything


Quick Reference

The 5 System Review Operations

OperationFocusTimeOutput
Ecosystem HealthOverall quality, completeness1-2hHealth score, strengths/weaknesses
Toolkit EffectivenessTool utilization, efficiency1-1.5hTool assessment, utilization metrics
Process EfficiencyDevelopment process optimization1-2hBottlenecks, automation opportunities
Coverage GapsMissing capabilities45-90mGap analysis, prioritized builds
System OptimizationSystem-level improvements1-2hRecommendations, action plan

Total Time: 5-8 hours for complete system review

System vs Individual Skill Reviews

system-reviewer (this skill):

  • Reviews the ECOSYSTEM
  • Systemic issues and patterns
  • System-level optimization
  • Meta-level assessment

review-multi:

  • Reviews INDIVIDUAL SKILLS
  • Skill-specific issues
  • Individual skill improvements
  • Skill-level assessment

Use Both: system-reviewer for ecosystem health, review-multi for individual quality

Integration with Continuous Improvement

system-reviewer → Identify systemic issues
    ↓
process-optimizer → Optimize processes
    ↓
feedback-analyzer → Analyze effectiveness
    ↓
auto-updater → Apply system improvements
    ↓
evolution-reporter → Report progress
    ↓
Improved ecosystem → Better skills

system-reviewer enables meta-level ecosystem assessment and system optimization for continuous ecosystem improvement.

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

github-copilot

28.23%
按下载量换算35

Claude Code

22.68%
按下载量换算28

mcpjam

17.1%
按下载量换算21

moltbot

10.66%
按下载量换算13

windsurf

6.89%
按下载量换算9

zencoder

3.46%
按下载量换算4

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。

来源信息

继续浏览同类 Skills