ClawHub

Skill Curator

Security checks across malware telemetry and agentic risk

Overview

This skill has a coherent curator purpose, but it can automatically turn Discord links into persistent GitHub skill changes without documented review or authorization controls.

Install only if you control the Discord trigger surface and the target GitHub repository. Use a fine-grained token limited to the intended repo, prefer pull requests or manual review before pushes, sanitize keywords into safe paths, and treat all extracted URL content as untrusted before it becomes part of any SKILL.md file.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill description does not clearly warn users that submitting a message will cause the system to fetch external URLs, transform the content, write files into a repository, and push changes to GitHub automatically. In this context, that omission is security-relevant because the skill performs network access and repository modification based on user input, which can lead to unexpected disclosure, persistence of untrusted content, or unintended commits.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill documents use of third-party analysis services and privileged credentials but does not include a privacy/security notice explaining that user-supplied content may be sent to external providers and that repository credentials are involved. This is dangerous because users may unknowingly submit sensitive URLs or content that gets transmitted off-platform and persisted, while operators may underestimate the trust boundary introduced by external services.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal