The Agent Skills Directory

[PROMPT_INJECTION]: The skill contains standard instructional guidelines and does not attempt to override agent safety or bypass instructions. It emphasizes reliability and data-driven analysis.\n- [DATA_EXFILTRATION]: No evidence of sensitive data exfiltration was found. Network activity is confined to well-known financial data providers (Yahoo Finance and crypto exchanges via CCXT), which are recognized as safe per trust rules.\n- [REMOTE_CODE_EXECUTION]: No remote code execution or untrusted software installations were detected. The skill relies on local scripts and standard, versioned dependencies from reputable package registries.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests external financial market data (OHLCV). Ingestion occurs via yfinance and ccxt in scripts/ta.py. Boundary markers include strict instructions to prioritize script output over natural language processing. Capabilities are limited to indicator computation and structured text output. Data is sanitized through conversion to numeric types and predefined labels, minimizing the injection surface.

analyse-technical