ClawHub

Bland Ai

Security checks across malware telemetry and agentic risk

Overview

The skill appears to be a real Bland AI integration, but it gives agents access to billable and outbound calling actions without clear confirmation boundaries.

Install only if you intend to connect Membrane to your Bland AI account. Treat read-only actions differently from actions that buy numbers, place calls, create batches, stop activity, or create external API tools, and require explicit user approval before any of those high-impact actions. Review and revoke the Membrane connection when it is no longer needed.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
82% confidence
Finding
The invocation text is overly broad ('Use when the user wants to interact with Bland AI data'), which can cause an agent to select this skill in situations involving sensitive or high-impact operations without sufficient user intent verification. Because the skill exposes actions that can place calls, purchase numbers, and modify workflows, ambiguous triggering increases the risk of unintended execution.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill enumerates high-impact actions such as purchasing phone numbers, sending calls, stopping calls, stopping batches, and creating operational resources, but provides no user-facing warnings or confirmation requirements. In an agent context, this can lead to real-world financial cost, service disruption, or unauthorized outbound communications if the skill is invoked carelessly or through prompt ambiguity.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal