可信 — 风险评分 5/100
上次扫描:1 天前 重新扫描
5 /100
complex-task-orchestrator
复杂任务编排与分治。当任务涉及批量操作、多步骤工程、sub-agent可能翻车的场景时激活。提供预记录防崩溃丢失、分治策略、超时管控、上下文膨胀防护、失败恢复方案。
The skill is a legitimate complex-task orchestration utility for AI sub-agent coordination. No malicious behavior, credential theft, obfuscation, or undeclared capabilities detected.
技能名称complex-task-orchestrator
分析耗时28.5s
引擎pi
可以安装
No action needed. The skill is safe to use.
资源类型声明权限推断权限状态证据
文件系统 WRITE WRITE ✓ 一致 SKILL.md: checkpoint file writes to memory/YYYY-MM-DD.md and checkpoints directo…
网络访问 NONE NONE No network requests in any file
命令执行 NONE NONE No subprocess, exec, or shell command execution
环境变量 NONE NONE checkpoint.py only uses os.environ for CHECKPOINT_DIR construction, no sensitive…
剪贴板 NONE NONE No clipboard access
浏览器 NONE NONE No browser access
数据库 NONE NONE No database access
技能调用 NONE NONE Only references other skills (company-research, excel-xlsx) declaratively in doc…

目录结构

3 文件 · 21.8 KB · 577 行
Markdown 2f · 465L Python 1f · 112L
├─ 📁 references
│ └─ 📝 common-patterns.md Markdown 143L · 4.1 KB
├─ 📁 scripts
│ └─ 🐍 checkpoint.py Python 112L · 3.8 KB
└─ 📝 SKILL.md Markdown 322L · 13.8 KB

安全亮点

✓ SKILL.md provides comprehensive, well-documented task orchestration guidance with no hidden capabilities
✓ checkpoint.py is a simple, readable utility (112 lines) with no external dependencies, no obfuscation, and no credential access
✓ Doc-to-code alignment: declared filesystem WRITE (checkpoints) matches actual implementation
✓ No network requests, no shell execution, no environment variable harvesting
✓ No base64, no eval, no curl|bash, no IP addresses for exfiltration
✓ Checkpoint data stored locally in user directory only
✓ references/common-patterns.md is pure documentation with no executable content
✓ No supply chain risk: checkpoint.py has no external dependencies