stock-analyser
Pass
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
curlvia the Bash tool to fetch real-time financial data from the official NSE India website. This is a legitimate and documented use case for stock analysis. The instructions include specific headers and cookie management required to navigate the platform's bot protection. - [EXTERNAL_DOWNLOADS]: The skill fetches data from well-known financial platforms including NSE India, Screener.in, Tickertape, and Moneycontrol. All URLs follow predictable, non-obfuscated patterns and target established financial services.
- [REMOTE_CODE_EXECUTION]: While the skill orchestrates parallel agents to collect data, these agents are used for structured data retrieval and analysis. There are no patterns suggesting the execution of untrusted remote scripts or arbitrary code injection.
- [DATA_EXFILTRATION]: No exfiltration patterns were detected. The skill reads public financial data and writes the final report to a local Markdown file. It does not access sensitive system files, environment variables (like SSH or AWS keys), or send user data to unauthorized external servers.
- [PROMPT_INJECTION]: The instructions are robust and include clear 'Non-negotiable rules' that prevent the AI from generating speculative content, 'multibagger' hype, or financial advice. There are no attempts to bypass safety filters or override system instructions.
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external websites (Screener.in, news articles, etc.), which is an inherent surface for indirect injection. However, the risk is mitigated by the skill's strict 'Tables first, prose second' requirement and the factual, data-centric nature of the task. The severity is assessed as LOW per standard guidelines.
Audit Metadata