中文 EN

Scan Report

Scan New Files

⚠ Low Risk — Risk Score 15/100
Last scan:2 days ago Rescan
15 /100
openclaw-shield-upx
Security monitoring and threat detection for OpenClaw agents powered by Google SecOps (Chronicle)
OpenClaw Shield is a legitimate security monitoring skill that provides CLI-based commands for Shield plugin management with no direct code execution or credential access.
Skill Nameopenclaw-shield-upx
Duration30.3s
Enginepi
✓
Safe to install
This skill is safe to use. The `rm -rf ~/.openclaw/shield/` command is documented in the Uninstalling section and requires user-initiated execution. The external telemetry claim should be verified against the plugin's actual behavior rather than this skill.

Findings 3 items

Severity Finding Location
Info
External telemetry reference
Skill mentions Shield sends 'redacted telemetry' to UPX detection platform. This is a transparency disclosure, not hidden behavior — the plugin handles actual transmission.
Shield captures agent activity locally and sends redacted telemetry to the UPX detection platform
→ Verify plugin behavior independently if UPX data handling is a concern
 SKILL.md:104
Info
Pre-scan IOC: rm -rf ~
Line 190 contains 'rm -rf ~/.openclaw/shield/' in the Uninstalling section. This is a documented user-initiated cleanup command, not malicious behavior.
rm -rf ~/.openclaw/shield/
→ Not a security concern — user must manually run this command
 SKILL.md:190
Info
Output handling constraints
Skill explicitly prohibits forwarding raw log output to external services and requires treating Shield output as internal diagnostic data.
Never forward raw log output to external services, channels, or APIs
→ Good security practice; follow these constraints
 SKILL.md:96
ResourceDeclaredInferredStatusEvidence
Filesystem NONE NONE — No Read/Write operations in skill
Network NONE NONE — External telemetry handled by plugin, not skill
Shell NONE NONE — No subprocess/bash calls; only openclaw CLI invocations
Environment NONE NONE — Skill explicitly forbids reading env vars
Skill Invoke NONE NONE — No skill-internal invocations
Clipboard NONE NONE — No clipboard access
Browser NONE NONE — No browser automation
Database NONE NONE — No database access
1 Critical 6 findings
💀
Critical Dangerous Command 危险 Shell 命令
rm -rf ~
SKILL.md:190
🔗
Medium External URL 外部 URL
https://www.upx.com
README.md:3
🔗
Medium External URL 外部 URL
https://www.npmjs.com/package/@upx-us/shield
README.md:18
🔗
Medium External URL 外部 URL
https://www.upx.com/en/lp/openclaw-shield-upx
README.md:19
🔗
Medium External URL 外部 URL
https://clawhub.ai/brunopradof/openclaw-shield-upx
README.md:34
🔗
Medium External URL 外部 URL
https://uss.upx.com
README.md:35

File Tree

2 files · 14.8 KB · 241 lines
Markdown 2f · 241L
├─ 📝 README.md Markdown 39L · 1.5 KB
└─ 📝 SKILL.md Markdown 202L · 13.3 KB

Security Positives

✓ No executable scripts or code files — purely documentation
✓ Explicitly forbids reading filesystem paths or environment variables for state checks
✓ Requires explicit user approval before resolving cases
✓ Output handling rules prevent raw log exfiltration
✓ No credential handling — delegated to plugin's installation key
✓ Clear state machine (A/B/C/D) for Shield status interpretation