Security audit

股票筛选入门工具

Security checks across malware telemetry and agentic risk

Overview

The skill is mostly a stock-screening helper, but its broad activation wording and inconsistent export/network behavior make it worth manual review before installation.

Install only if you are comfortable with an agent using Python execution and external AkShare/public market-data services for stock screening. Treat the free edition as terminal-display only unless the publisher clarifies the export contradiction, and avoid allowing this skill to handle unrelated data-analysis tasks.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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 (5)

Description-Behavior Mismatch

Medium

Confidence: 93% confidence
Finding: The skill documentation says the free edition does not support export, but the core execution section explicitly states that `input_params` supports create/query/export operations. This inconsistency can cause an agent to expose or attempt data export paths that users and reviewers would reasonably believe are unavailable, weakening safety expectations and permission boundaries.

Intent-Code Divergence

Medium

Confidence: 90% confidence
Finding: The documentation contradicts itself about export support, which creates ambiguity in what the agent is allowed to do. In agent settings, contradictory capability claims are dangerous because policy enforcement, user expectations, and downstream automation may rely on the documented edition limits.

Vague Triggers

High

Confidence: 95% confidence
Finding: The activation scope is overly broad and includes generic phrases like data analysis, report generation, statistical insight, and visualization, which are not narrowly tied to stock screening. This can cause the skill to trigger in unrelated contexts, increasing the chance that an `exec`-capable finance skill is selected for tasks outside its intended domain and gains unnecessary access to user workflows or data.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The repeated 'Use when' wording is ambiguous and lacks clear boundaries for when the skill should or should not be invoked. Ambiguous routing text increases the risk of accidental activation, especially because the skill declares `read`, `write`, and `exec` tools and may perform network-backed operations.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The skill documents external data retrieval through AkShare and public data sources but does not clearly warn users that running it may access remote services. In an agent environment, undisclosed network use can leak prompts, symbols, or analysis parameters to third parties and may violate user expectations or enterprise policy.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.