ClawHub

Faithful Task Executor

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed task-planning workflow skill that stores task rules locally and coordinates subagents, with no evidence of hidden code, credential access, exfiltration, or destructive behavior.

Install only if you want a structured task workflow that may activate on broad task-execution phrases. Review the generated .task-rules/rules.md before confirming it, avoid placing secrets in task rules, and separately review any scripts if a later version adds the referenced scripts directory.

SkillSpector

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

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger phrase "执行这个任务" is extremely generic and can match ordinary user requests, which creates ambiguous invocation boundaries for the skill. In an agent ecosystem, this can cause the skill to activate unexpectedly and impose its workflow, rule-locking, and subagent orchestration on unrelated tasks, increasing the chance of prompt-routing errors or unintended control over task execution.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The activation phrases are generic everyday requests such as '执行这个任务', '处理这个工作', and '完成以下要求', so the skill can trigger during ordinary conversations that were not intended to invoke this control layer. Because this skill changes execution behavior by enforcing rule persistence and sub-agent coordination, unintended activation can alter downstream task handling, create hidden workflow side effects, and increase the chance that other safeguards or user expectations are bypassed.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal