Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

EdgeIQ XSS Scanner

v1.0.0

Scans web apps for reflected and DOM-based XSS using 24+ payloads across 6 contexts, with crawl and concurrency support for authorized security audits.

0· 31·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Suspicious
medium confidence
Purpose & Capability
Name/description (XSS scanning) match the code: scanner.py implements payload injection, crawling, and reflection detection in pure Python. However, there are mismatches in the documentation vs code (e.g., SKILL.md mentions 'full port coverage' in an example but scanner.py contains no port-scanning logic). The discord wrapper and install instructions also assume particular filesystem layout that doesn't align with the installation advice.
Instruction Scope
SKILL.md instructs only running the scanner against authorized targets and shows CLI usage; the runtime instructions are consistent with network scanning behavior. The discord wrapper runs the scanner via subprocess and returns its stdout/stderr to the caller (expected for a CLI wrapper). Minor scope issues: the wrapper hardcodes a path (/home/guy/.openclaw/workspace/apps/xss-scanner/scanner.py) and the HELP text defaults (workers=15) differ from SKILL.md defaults (workers=5). These are coherence/usability concerns rather than direct exfiltration instructions; the scanner does perform arbitrary HTTP requests to targets you supply.
Install Mechanism
No install spec (instruction-only) and the code is pure stdlib Python — low install risk. The SKILL.md includes a manual cp instruction which assumes a local source path; nothing is downloaded from remote URLs or executed during install.
Credentials
The skill requests no environment variables, no credentials, and no config paths. This is proportionate to its stated purpose. The code does create an SSL context that disables certificate verification (verify_mode = CERT_NONE), which is typical for scanners that need to hit self-signed hosts but reduces TLS safety — not a credential leak but a security-quality note.
Persistence & Privilege
The skill does not request persistent/always-on privileges and is user-invocable only. It does spawn subprocesses to run scanner.py, which is expected. There is no code modifying other skills or agent-wide settings.
What to consider before installing
This skill appears to be a straightforward, pure-Python XSS scanner, but review the following before installing or running: - Hardcoded path: discord_xss_command.py expects scanner.py at /home/guy/.openclaw/workspace/apps/xss-scanner/scanner.py. Update the path or place files accordingly to avoid surprises. - Documentation mismatches: SKILL.md examples mention 'full port coverage' and default workers=5, whereas the wrapper help mentions workers=15 and the code contains no port-scanning. Treat these as sloppy documentation and verify functionality matches your expectations before use. - TLS behavior: scanner.py disables certificate verification (CERT_NONE). If you need strict TLS checks, modify build_ssl_context. - Network impact & legality: the tool issues HTTP requests to targets you provide (and can follow external links if enabled). Only scan systems you own or have explicit permission to test. Consider running scans from an isolated environment or lab to avoid collateral impact. - Output and disclosure: results are returned to the caller (and the discord wrapper returns stdout/stderr). Confirm where results will be sent/posted (Discord channel, logs) to avoid leaking sensitive findings. If you plan to use this skill: (1) correct the hardcoded path or inspect the wrapper to ensure it calls the intended scanner, (2) run the scanner locally against a known test target first, (3) consider a quick audit of the rest of scanner.py (the provided code is readable) to confirm no hidden telemetry or external reporting, and (4) only run against authorized targets.

Like a lobster shell, security has layers — review code before you run it.

auditingvk978v59s6787e1qhz4zpc7ddfn84td28latestvk978v59s6787e1qhz4zpc7ddfn84td28securityvk978v59s6787e1qhz4zpc7ddfn84td28vulnerability-scanvk978v59s6787e1qhz4zpc7ddfn84td28web-securityvk978v59s6787e1qhz4zpc7ddfn84td28xssvk978v59s6787e1qhz4zpc7ddfn84td28

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Comments