ClawHub

Api Tester From Bilibili

Security checks across malware telemetry and agentic risk

Overview

This is a non-executable API-testing skill with noisy auto-merged content, but no evidence of malicious actions or hidden access.

Install only if you are comfortable with an auto-generated skill that may route on broad API-testing phrases and contains irrelevant learned notes. It appears safe from an execution and access standpoint, but its content should be cleaned up for reliability.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Intent-Code Divergence

Medium
Confidence
93% confidence
Finding
The skill embeds unrelated and contradictory material, including train-spotting and consumer-product content, inside an API-testing skill. In an agent setting this can corrupt routing, confuse downstream prompt selection, and cause the model to execute or recommend irrelevant actions, which weakens reliability and can become a prompt-injection-style control problem when arbitrary learned content is merged into operational instructions.

Vague Triggers

Medium
Confidence
88% confidence
Finding
The trigger phrases are broad everyday terms for API testing, so the skill may activate in ordinary conversation without clear user intent. Over-broad activation increases the chance of unintended prompt takeover or tool routing, especially because the skill content already contains noisy unrelated instructions that could steer behavior once invoked.

Vague Triggers

Medium
Confidence
84% confidence
Finding
The manifest trigger value "api tester" is generic and likely to overlap with normal user discussion, making accidental activation plausible. By itself this is not severe, but in the context of an auto-generated skill with contaminated content it increases exposure to misrouting and unintended instruction influence.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal