"i问财选股技能"

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:19 PM UTC
• Version:
  1.0.0

Overview

This is a coherent stock-screening helper that queries i问财 and creates an Excel report, with no evidence of hidden credential use, destructive actions, or unrelated data access.

Install only if you are comfortable with stock-related terms from your links, files, text, or keywords being used for i问财 searches and with a local Excel/CSV report being created. Do not provide confidential documents or private investment material unless you are comfortable with derived query terms leaving the local environment.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (6)

Vague Triggers

Medium

• Confidence: 93% confidence
• Finding: The trigger conditions are broad phrases such as 'XX选股', '根据XX选股', and generic link/text-plus-analysis requests, which can cause the skill to activate in contexts the user did not clearly intend. In an agent environment, overly permissive activation can lead to unsolicited browsing of i问财, processing of arbitrary user-provided links or files, and unintended generation of stock-selection outputs and artifacts.

Missing User Warnings

Medium

• Confidence: 84% confidence
• Finding: The README states that the skill generates Excel output but does not warn users that it will write a file to disk. In agent or shared execution environments, silent file creation can surprise users, leak sensitive derived data into persistent storage, or overwrite/accumulate artifacts without explicit consent.

Vague Triggers

High

• Confidence: 92% confidence
• Finding: Allowing the model to 'self-judge' semantically when to activate the skill creates ambiguous routing boundaries and can cause the skill to run on loosely related requests. In this context, that increases the chance of unnecessary external access, unintended local file creation, and inappropriate use of a finance-related workflow without clear user intent.

Vague Triggers

Medium

• Confidence: 85% confidence
• Finding: Broad phrases such as '根据XX选股' and generic requests about analyzing beneficial industries overlap with ordinary analytical conversations, so the skill may activate when the user only wanted discussion rather than stock screening actions. In a stock-selection skill, that can lead to unintended downstream behavior, including external processing and filesystem writes, without sufficiently specific consent.

Missing User Warnings

Medium

• Confidence: 88% confidence
• Finding: The skill instructs generation of an Excel file at a local filesystem path but does not clearly warn users in the description that local files will be created. Hidden persistence is risky because it creates side effects on the host environment, may overwrite or expose sensitive data, and prevents informed consent about local storage.

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The skill says it will process user-provided links, files, or text and use an external website for stock queries, but it does not disclose the privacy implications of transmitting or deriving information from that content externally. If users provide proprietary documents, private links, or sensitive text, the lack of warning can result in unintentional data exposure to third parties.

VirusTotal

65/65 vendors flagged this skill as clean.