Low Risk — Risk Score 10/100
Last scan:2 days ago Rescan
10 /100
smart-model-switcher-v5
智能模型自动切换 V5.0.2 - 多模态感知,自动识别图片/视频/音频/代码/文本任务
This is a legitimate AI model routing skill that automatically switches between different AI models based on task type detection. No malicious behavior detected.
Skill Namesmart-model-switcher-v5
Duration25.3s
Enginepi
Safe to install
This skill is safe to use. No security concerns requiring action.

Findings 2 items

Severity Finding Location
Low
Non-executable script format
PowerShell scripts in scripts/ directory have .txt extension instead of .ps1, suggesting they are documentation rather than executable code
auto-monitor.txt, runtime-switch.txt
→ Confirm if scripts need to be converted to .ps1 for actual execution
scripts/
Low
Encoding issues in Chinese keywords
Chinese keywords in runtime-switch.txt appear garbled, likely due to encoding problems during file creation
$TaskKeywords - contains corrupted Chinese text
→ Re-encode the keyword arrays with proper UTF-8 encoding
scripts/runtime-switch.txt:35
ResourceDeclaredInferredStatusEvidence
Filesystem NONE NONE No filesystem access detected in scripts
Network NONE NONE No network calls in code
Shell NONE NONE PowerShell scripts are text descriptions only, not executable
Environment NONE NONE No environment variable access
Skill Invoke NONE NONE SKILL.md mentions session_status tool but no actual invocation in scripts

File Tree

3 files · 26.2 KB · 752 lines
Markdown 1f · 484L Text 2f · 268L
├─ 📁 scripts
│ ├─ 📄 auto-monitor.txt Text 90L · 3.4 KB
│ └─ 📄 runtime-switch.txt Text 178L · 6.4 KB
└─ 📝 SKILL.md Markdown 484L · 16.3 KB

Security Positives

✓ No credential harvesting or exfiltration code present
✓ No external network requests detected
✓ No base64 encoding or obfuscated commands
✓ No suspicious system path access (~/.ssh, ~/.aws, etc.)
✓ No reverse shell or C2 indicators
✓ Skill functionality is clearly documented in SKILL.md
✓ Task routing logic is straightforward and transparent