算卦（小六壬 · 大六壬 · 六爻 · 梅花易数）

• Outcome: Pass
• Latest audit: Jun 16, 2026 at 3:54 PM UTC
• Version:
  1.0.1

Overview

This divination skill is coherent and disclosed, with the main user considerations being optional city geocoding, local caching, and broad trigger words.

Install only if you are comfortable with a divination tool that may send unknown city names to Open-Meteo and cache resolved city coordinates locally. For maximum privacy, use the explicit longitude option or known built-in cities, and treat any outputs as entertainment or cultural reference rather than advice for medical, legal, financial, or other major decisions.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands

Findings (5)

Description-Behavior Mismatch

Medium

• Confidence: 93% confidence
• Finding: The skill is presented as a local divination/calculation utility, but this code adds undeclared external geocoding and writes a persistent local cache. That creates privacy and trust issues because user-supplied location data may leave the system and be stored on disk even when the user likely expects an offline calculation workflow.

Context-Inappropriate Capability

Medium

• Confidence: 88% confidence
• Finding: This function sends user-provided city names to an external geocoding API, expanding the skill's capability beyond pure local divination logic. Even if used to improve true-solar-time calculations, the external call introduces data exfiltration, third-party dependency, and availability risks that are not clearly constrained by the stated purpose.

Vague Triggers

Medium

• Confidence: 88% confidence
• Finding: The trigger list includes very broad everyday Chinese terms such as '算', '算算', and '预测', which can appear in many non-divination contexts. In an agent environment, this raises the chance of unintended skill activation, causing the agent to route unrelated user requests into this skill and potentially perform unexpected network calls or misleading fortune-telling behavior.

Vague Triggers

Low

• Confidence: 78% confidence
• Finding: The architecture example uses the phrase '帮我算个事', which is conversationally vague and can overlap with ordinary requests about calculation, planning, or advice. If implementers mirror this example as a trigger boundary, the skill may activate on ambiguous language and produce irrelevant or confusing outputs.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: At this call site, user input is passed into external geocoding without any warning or consent prompt. A user asking a divination question may not expect their location string to be transmitted to a third party, so this is a privacy vulnerability caused by insufficient transparency and consent.

VirusTotal

65/65 vendors flagged this skill as clean.