ClawHub
Back to skill

Security audit

Prism Scanner

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward crypto-token scanner that calls a disclosed PRISM API, with privacy and trigger-word caveats but no evidence of hidden or harmful behavior.

Install only if you are comfortable sending token symbols or contract addresses to the configured PRISM API. Treat results as a screening signal, not financial advice or a complete independent audit, and prefer explicit invocation wording to avoid accidental scans.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
89% confidence
Finding
The skill advertises shell-based usage (`./scan.sh ...`) but declares no permissions, creating a capability/permission mismatch. In agent environments, undeclared shell execution can lead to unexpected command execution paths, reduced operator visibility, and weaker policy enforcement around what the skill is allowed to do.

Tp4

High
Category
MCP Tool Poisoning
Confidence
84% confidence
Finding
The skill description presents deterministic local rug-pull checks, but the documented behavior depends heavily on an external remote analysis service and includes additional functionality such as copycat detection. This mismatch is dangerous because users and agents may trust the skill to perform transparent, specific checks when it is actually delegating trust to a third-party service whose logic, data handling, and outputs are not visible or auditable from the skill itself.

Vague Triggers

Medium
Confidence
92% confidence
Finding
The trigger phrases are generic natural-language patterns such as "check {token}" and "analyze {token}", which can collide with normal user requests and cause the skill to activate unexpectedly. In an agent environment, this can route unrelated prompts into this skill, leading to incorrect execution context, unintended external API calls, and confusing or unsafe automation behavior.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.