Trusted — Risk Score 5/100
Last scan:2 days ago Rescan
5 /100
breakfast-recommender
智能早餐推荐助手,根据冰箱食材推荐早餐、管理食材、记录偏好与历史
A straightforward breakfast recommendation skill with no security concerns—operates solely on filesystem operations to a well-defined directory for managing fridge inventory and history.
Skill Namebreakfast-recommender
Duration101.5s
Enginepi
Safe to install
This skill is safe to use. No security controls needed.

Findings 1 items

Severity Finding Location
Info
Sample data files present
The references/ directory contains sample fridge.md and history.md files with hardcoded test data (e.g., eggs, hand pie, corn). These are legitimate sample data for demonstration purposes.
鸡蛋(数量未知), 手抓饼 x 2, 玉米 x 1 根
→ No action needed. Sample data is clearly labeled in the references/ directory and serves as documentation.
references/fridge.md, references/history.md:1
ResourceDeclaredInferredStatusEvidence
Filesystem WRITE WRITE ✓ Aligned SKILL.md declares writes to ~/.openclaw/workspace/breakfast-recommender/
Shell WRITE NONE ✓ Aligned cron usage declared in Scenario 3, no other shell execution observed
Network NONE NONE No network activity detected
Environment NONE NONE No environment variable access observed
Database NONE NONE Uses flat markdown files, not a database

File Tree

3 files · 6.8 KB · 189 lines
Markdown 3f · 189L
├─ 📁 references
│ ├─ 📝 fridge.md Markdown 12L · 150 B
│ └─ 📝 history.md Markdown 6L · 281 B
└─ 📝 SKILL.md Markdown 171L · 6.3 KB

Security Positives

✓ All file operations are scoped to a well-defined directory: ~/.openclaw/workspace/breakfast-recommender/
✓ No network requests or external communications
✓ No credential, SSH key, or sensitive path access
✓ No shell command injection vectors (eval, base64, subprocess to external commands)
✓ Cron usage for scheduling is explicitly declared in SKILL.md
✓ Data format is simple markdown—no binary payloads or encoded content
✓ No third-party dependencies or package requirements
✓ Skill is purely a lifestyle tool with no attack surface