T09 · Insecure Skill Coding Practices
- Location
scripts/fetch_stock.py:113- Finding
Unauthenticated HTTP Allows Manipulation of Financial Market Data
- Content
View full analysis
- Remediation
View remediation
Security audit
Security checks for vulnerabilities and agentic risk
This A-share stock assistant is not clearly malicious, but it warrants review because it gives actionable trading guidance from unauthenticated market data and stores user alert/watchlist information.
Before installing, treat this as an unregulated market-analysis helper rather than a financial adviser. Be aware that it may save your alert/watchlist interests locally, and verify prices and any buy/sell or position-sizing suggestions against a trusted brokerage or market-data source before acting.
scripts/fetch_stock.py:113Unauthenticated HTTP Allows Manipulation of Financial Market Data
The trigger conditions are written in mixed Chinese and English and rely on descriptive phrases rather than precise boundaries, making invocation ambiguous. This ambiguity can lead to accidental activation or inconsistent behavior across user inputs, which is risky for a skill that claims real-time market data access and can influence trading decisions.
The skill instructs persisting user alert settings to a local file without an explicit user-facing notice or consent flow. Even if the stored data seems limited, watchlist entries and price alerts reflect user investment interests and behavior, so silent retention creates a privacy risk and may expose sensitive financial preferences to other processes or future sessions.
The document gives concrete buy/sell, position-sizing, and market-timing guidance such as '可做多', '买入信号', '卖出信号', and recommended仓位, but it does not include any warning that the material is educational or that investing involves risk. In an agent skill context, users may reasonably treat this as actionable financial advice, increasing the chance of harmful decisions and potential compliance or trust issues.
The trigger conditions are written in mixed Chinese and English and rely on descriptive phrases rather than precise boundaries, making invocation ambiguous. This ambiguity can lead to accidental activation or inconsistent behavior across user inputs, which is risky for a skill that claims real-time market data access and can influence trading decisions.
The natural-language content forces a specific language/locale throughout the file. The policy allows locale constraints only when the skill offers user choice or clearly documents a justified regional limitation, neither of which appears here.
This markdown file contains user-facing instructions exclusively in Chinese and does not state that the language is optional or region-specific. Under the language/locale policy, forcing a specific language without user opt-in can be a natural-language policy violation.
The request header hard-codes "Accept-Language: zh-CN,zh;q=0.9", which imposes a specific locale preference. Under the policy, language or locale constraints should be optional or clearly justified; this file does not offer user opt-in or explain why Chinese locale is required.