可信 — 风险评分 0/100
上次扫描:2 天前 重新扫描
0 /100
cjl-autoresearch-cc
Improve skills, prompts, articles, workflows, and systems via iterative mutation-testing
This autoresearch skill is a safe, read-only content improvement tool that only reads files and performs in-memory text mutations without any network, shell, or credential access.
技能名称cjl-autoresearch-cc
分析耗时25.7s
引擎pi
可以安装
This skill is safe to use. No security concerns identified.
资源类型声明权限推断权限状态证据
文件系统 READ READ ✓ 一致 Skill reads target SKILL.md/plugin files for analysis only
网络访问 NONE NONE No network operations described or used
命令执行 NONE NONE No shell commands, subprocess, or eval patterns
环境变量 NONE NONE No environment variable access
技能调用 READ READ ✓ 一致 Can read other skills for optimization purposes
剪贴板 NONE NONE No clipboard access
浏览器 NONE NONE No browser operations
数据库 NONE NONE No database access

目录结构

5 文件 · 35.6 KB · 1175 行
Markdown 5f · 1175L
├─ 📁 references
│ └─ 📝 mutation_strategies.md Markdown 114L · 3.7 KB
├─ 📝 CLAUDE.md Markdown 9L · 216 B
├─ 📝 README_zh.md Markdown 377L · 9.9 KB
├─ 📝 README.md Markdown 377L · 10.4 KB
└─ 📝 SKILL.md Markdown 298L · 11.4 KB

安全亮点

✓ Pure text-processing skill with no code execution capabilities
✓ All operations are in-memory with no disk writes (test cases stored in context)
✓ Documentation accurately describes behavior - no hidden functionality
✓ No external dependencies or package requirements
✓ Target file access is scoped to specific ~/.claude/skills/ and plugin paths
✓ No credential, SSH key, or sensitive file access attempted
✓ No network requests or data exfiltration vectors
✓ No obfuscation techniques (base64, eval, etc.)
✓ Mutation strategies are document-centric (text quality improvements only)