Low Risk — Risk Score 10/100
Last scan:2 days ago Rescan
10 /100
seedream5.0
使用 Seedream5.0 接口进行文生图与参考图生成
Legitimate image generation skill with properly declared API key storage, environment variable access, and external network calls. No malicious behavior detected.
Skill Nameseedream5.0
Duration24.0s
Enginepi
Safe to install
Skill is safe for use. Monitor external API endpoint availability and ensure X_API_KEY is obtained from the legitimate source kexiangai.com.

Findings 2 items

Severity Finding Location
Low
External API endpoint dependency
Skill depends on external service at agent.mathmind.cn. Service availability or security is outside skill's control.
https://agent.mathmind.cn/minimalist/api/volcengine/ai/fzGenerateImg5
→ Verify endpoint reliability or implement fallback handling
SKILL.md:68
Low
API key source externalization
Users directed to kexiangai.com for API key - this external site is not verified by the skill.
x-api-key(kexiangai.com)
→ Ensure API key source is legitimate before use
SKILL.md:13
ResourceDeclaredInferredStatusEvidence
Filesystem WRITE WRITE ✓ Aligned SKILL.md:119 writes to ~/.config/seedream5.0/.env
Environment READ READ ✓ Aligned SKILL.md:48 reads X_API_KEY env var
Network WRITE WRITE ✓ Aligned SKILL.md:68 POSTs to external API
Shell WRITE WRITE ✓ Aligned scripts/generate.sh and set_key.sh use bash for API calls
1 findings
🔗
Medium External URL 外部 URL
https://agent.mathmind.cn/minimalist/api/volcengine/ai/fzGenerateImg5
SKILL.md:68

File Tree

4 files · 8.6 KB · 302 lines
Markdown 2f · 170L Shell 2f · 132L
├─ 📁 references
│ └─ 📝 api-guide.md Markdown 34L · 982 B
├─ 📁 scripts
│ ├─ 🔧 generate.sh Shell 113L · 2.7 KB
│ └─ 🔧 set_key.sh Shell 19L · 322 B
└─ 📝 SKILL.md Markdown 136L · 4.6 KB

Security Positives

✓ API key is masked in logs (SKILL.md:21: '日志与回显仅允许掩码展示')
✓ File permissions properly set to 600 for credential storage (scripts/set_key.sh:18)
✓ Network timeout configured (scripts/generate.sh:9: TIMEOUT=600)
✓ Input validation present (prompt required, size/watermark normalized)
✓ All capabilities declared in SKILL.md
✓ No hidden functionality or suspicious patterns detected
✓ No base64 encoding, eval(), or remote script execution
✓ No credential exfiltration or suspicious network destinations