ClawHub

HR面试评价助手

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only HR evaluation skill that handles sensitive candidate documents for its stated purpose, with no evidence of hidden code, exfiltration, persistence, or privileged actions.

Install only if you are authorized to process candidate resumes and interview notes in this environment. Treat generated scores and hire/no-hire recommendations as decision support requiring human review, and verify any external export scripts or dependencies before running them.

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
95% confidence
Finding
The skill auto-activates on very broad file keywords such as “简历”, “面试”, “candidate”, and “evaluation” combined with common document types. This can cause unintended invocation on unrelated uploads, leading the agent to parse sensitive HR documents or begin a workflow without clear user intent, which is especially risky because resumes and interview notes contain personal and confidential employment data.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill handles highly sensitive candidate data from resumes and interview records and also exports derived reports, but the user-facing flow does not clearly disclose what data will be extracted, retained, shared, or written into output files. This creates privacy and compliance risk because users may unknowingly process personal data, and exported PDF/PNG artifacts can propagate sensitive information beyond the original conversation.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal