Back to skill

Security audit

Stock Watcher

Security checks for vulnerabilities and agentic risk

Overview

This stock monitoring skill is disclosed and purpose-aligned, with financial-advice and always-on monitoring risks users should understand before use.

Review this as a financial monitoring and analysis aid, not professional investment advice. It may run as a long-lived monitor and handle your watchlist, cost basis, and portfolio configuration, so only provide data you are comfortable storing locally and using with the tool's market-data integrations.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The description states the skill is tailored to Chinese investors and uses Chinese-language content and China-specific market conventions. This imposes a locale/language constraint without an explicit opt-in or alternative language choice, which can violate language/locale policy requirements.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

The sample alert output includes '深度分析' and 'Kimi建议', and the file list documents modules for 'risk_advice.py' and detailed analysis tools. These features go beyond a straightforward monitoring and warning system and move into interpretive investment analysis/advice, which is not reflected in the manifest description.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest description focuses on a '股票监控预警系统' with seven alert rules for watched positions. However, the documented file structure lists modules such as 'pick_short_term.py', 'screen_stocks.py', and 'stock_analyzer.py', which imply active stock selection and strategy generation rather than merely monitoring and alerting on predefined securities.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The manifest highlights seven major alert rules as the skill's core functionality. The file description for 'monitor.py' says '7大规则 + 操盘手检测', introducing an additional behavioral analysis capability not represented in the manifest's stated scope.

Content

No source excerpt is available for this finding.