stock-trading
Pass
Audited by Gen Agent Trust Hub on Mar 21, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill makes network requests to external APIs for real-time market data. It connects to Sina Finance via
hq.sinajs.cnand to an OpenCTC simulation gateway at121.37.80.177. These are established services for financial data and trading simulation. - [PROMPT_INJECTION]: The skill's trading strategy involves ingesting data from web searches, which could contain adversarial instructions designed to influence the agent's behavior.
- Ingestion points: Untrusted data enters the context via search results as specified in the workflow instructions in
strategy.md. - Boundary markers: The instructions include a clear safety rule requiring explicit user confirmation before any transaction is executed, which serves to mitigate the risk of automated unauthorized actions.
- Capability inventory: The scripts perform file writes to a local
state.jsonfor portfolio tracking and use theurllibandopenctp_ctplibraries for network-based data retrieval across all scripts. - Sanitization: The skill relies on agent reasoning and the mandatory confirmation workflow rather than automated input filtering or escaping of the search results.
Audit Metadata