中文 EN

Scan Report

Scan New Files

⚠ Low Risk — Risk Score 25/100
Last scan:20 hr ago Rescan
25 /100
MeiHuaYi (梅花易数)
Traditional Chinese I Ching divination system with plum blossom numerology calculations
Legitimate divination system with undeclared network access for data download and undocumented filesystem WRITE permissions - no malicious behavior detected.
Skill NameMeiHuaYi (梅花易数)
Duration39.4s
Enginepi
✓
Safe to install
Add network:READ and filesystem:WRITE declarations to SKILL.md. Pin the iching.json source URL to a specific version/commit to prevent supply chain risks.

Findings 3 items

Severity Finding Location
Medium
Undeclared network access for data download Doc Mismatch
meihuayi.py automatically downloads iching.json from GitHub (john-walks-slow/open-iching) on first run. This network READ operation is not declared in SKILL.md capabilities.
urllib.request.urlretrieve(_ICHING_URL, _ICHING_PATH)
→ Declare network:READ capability if this behavior is intentional and necessary for the skill's core divination functionality.
 meihuayi.py:25
Low
External data dependency without version pinning Supply Chain
The iching.json is fetched from a dynamic GitHub URL without version/commit pinning. Future changes to the upstream repository could alter the I Ching data, affecting prediction accuracy.
_ICHING_URL = "https://raw.githubusercontent.com/john-walks-slow/open-iching/master/iching/iching.json"
→ Consider pinning to a specific commit hash or include iching.json directly in the skill package to eliminate external dependency.
 meihuayi.py:23
Low
Undocumented filesystem WRITE permissions Doc Mismatch
The skill creates data directories and SQLite databases (divination_records.db, learning_notes.db) in the data/ folder. This WRITE behavior is not declared in SKILL.md.
os.makedirs(_DATA_DIR); sqlite3.connect(_DB_PATH)
→ Declare filesystem:WRITE capability for SQLite database operations in data/ directory.
 records.py:17
ResourceDeclaredInferredStatusEvidence
Filesystem NONE WRITE ✗ Violation meihuayi.py:28-30 - os.makedirs(_DATA_DIR)
Network NONE READ ✗ Violation meihuayi.py:25 - urllib.request.urlretrieve from raw.githubusercontent.com
Shell NONE NONE — No subprocess or shell commands found
Environment NONE NONE — No os.environ access for sensitive keys
Skill Invoke NONE NONE — No inter-skill invocations
Clipboard NONE NONE — No clipboard access
Browser NONE NONE — No browser automation
Database NONE WRITE ✗ Violation records.py:17-18 - sqlite3.connect creates local DB in data/

File Tree

8 files · 140.5 KB · 5716 lines
JSON 1f · 3597L Markdown 4f · 1134L Python 3f · 985L
├─ ▾ 📁 data
│ ├─ 📋 iching.json JSON 3597L · 69.4 KB
│ ├─ 📝 万物类象.md Markdown 216L · 6.8 KB
│ ├─ 📝 三要十应.md Markdown 172L · 4.1 KB
│ └─ 📝 解卦技巧.md Markdown 212L · 8.4 KB
├─ 🐍 learning_notes.py Python 170L · 6.2 KB
├─ 🐍 meihuayi.py Python 393L · 17.5 KB
├─ 🐍 records.py Python 422L · 13.6 KB
└─ 📝 SKILL.md Markdown 534L · 14.5 KB

Dependencies 1 items

PackageVersionSourceKnown VulnsNotes
iching.json master (unpinned) https://raw.githubusercontent.com/john-walks-slow/open-iching/master/iching/iching.json No External JSON data fetched via urllib without version/commit pinning

Security Positives

✓ No credential harvesting - skill does not access ~/.ssh, ~/.aws, .env, or similar sensitive paths
✓ No data exfiltration - no network POST requests or external data transmission
✓ No shell execution - no subprocess, os.system, or command injection vectors
✓ No obfuscation - all code is readable Python without base64 or eval patterns
✓ No persistence mechanisms - no cron jobs, startup scripts, or backdoor installations
✓ No prompt injection - SKILL.md contains legitimate documentation without hidden instructions
✓ SQLite databases are stored locally in data/ directory, not sent externally