ClawHub

招中标信息大数据查询及分析-千里马

Security checks across malware telemetry and agentic risk

Overview

This is a coherent procurement-data search skill that uses a disclosed third-party API key, with privacy and scope cautions around company matching and contact lookups.

Install only if you trust the ZLBX/Qianlima service and are comfortable sending procurement queries and company names to its API. Keep the API key in environment or agent config storage, avoid confidential strategy queries unless approved, and ask the agent to confirm company matches or contact lookups when privacy or scope matters.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (3)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The description says the skill 'must' be invoked for broad categories like market analysis and aggregation queries, which can cause over-triggering outside the tool's narrow necessity. This increases the chance of unnecessary external data disclosure, unexpected API use, and agent routing errors when a simpler or safer local response would suffice.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The skill exposes a company contact lookup capability without any warning, consent check, or usage restriction around potentially sensitive personal or business contact information. In an agent setting, this can facilitate privacy-invasive enrichment, mass contact harvesting, or disclosure of personal data beyond user expectations.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill explicitly instructs automatic semantic matching of a company short name to multiple related legal entities and then performing follow-on queries across all matches without user confirmation. In a business-intelligence context this can cause over-broad data retrieval, mistaken aggregation across unrelated or only loosely related entities, and disclosure of contact or partner information the user did not clearly request.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal