ClawHub
Back to skill

Security audit

Privora · A股/港股/黄金/基金 多资产 量化分析 · 量化回测 · 模拟盘 · 实时告警 · 风险监控 · Python 策略 · AI Agent

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed Privora API wrapper, but its broad write/process/webhook authority and conflicting paper-trading token guidance need careful review before installation.

Install only with a dedicated low-scope token at first. Grant portfolio, process, scheduler, alert, webhook, or paper-trading scopes only for a specific workflow, and require human confirmation before state-changing calls or webhook triggers. Treat the paper-trading PAT instruction as unreliable unless the platform confirms the correct issuance flow.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
85% confidence
Finding
The skill exposes shell-based helper scripts and command examples but does not declare corresponding permissions, which can hide the true execution surface from users and policy engines. In an agent context, undeclared shell capability increases the chance that a host environment executes networked commands with bearer tokens or other sensitive environment variables without an explicit trust boundary.

Tp4

High
Category
MCP Tool Poisoning
Confidence
80% confidence
Finding
The documentation frames the skill as a domain-specific investment platform, but the described behavior includes a generic request-construction/dispatch wrapper, skill discovery, raw JSON pass-through, and dry-run features. That broader control surface makes the integration more powerful than the top-level description suggests, increasing the risk of over-trusting the skill and enabling unintended access to additional endpoints or request shapes.

Intent-Code Divergence

High
Confidence
92% confidence
Finding
The document first states that paper-trading scopes are reserved and not self-mintable, then later instructs users to create their own PAT with those scopes. Contradictory guidance around privileged token issuance is dangerous because it can cause operators or agents to attempt broader token minting patterns, misunderstand trust boundaries, and mishandle credentials for simulated trading and related data access.

External Transmission

Medium
Category
Data Exfiltration
Content
# {"id": 15, "assetName": "stock_dividend"}

# Step 2: 用数字 id (不是 assetName!) 查元数据
curl -H "Authorization: Bearer $LG_AGENT_TOKEN" \
  https://privora.cn/api/data-assets/42/metadata

# Step 3: 用数字 id 查实际数据
Confidence
78% confidence
Finding
curl -H "Authorization: Bearer $LG_AGENT_TOKEN" \ https://privora.cn/api/data-assets/42/metadata # Step 3: 用数字 id 查实际数据 curl -H "Authorization: Bearer $LG_AGENT_TOKEN" \ "https://privora.cn/api/d

VirusTotal

58/58 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.