The Agent Skills Directory

[COMMAND_EXECUTION]: The orchestration script scripts/pack.py uses subprocess.run to execute internal worker scripts (fdr_client.py and yfinance_client.py). This is a documented architectural pattern for the skill's multi-layered design. The commands are executed using uv run and the arguments are strictly managed via normalization functions (e.g., normalize_ticker), which reduces the risk of argument injection.
[EXTERNAL_DOWNLOADS]: The skill retrieves data from legitimate external sources to fulfill its stated purpose:
Fetches macroeconomic series from the Federal Reserve Bank of St. Louis (FRED) via the fred.stlouisfed.org domain.
Retrieves Korean equity price and financial data using the yfinance library.
Accesses Bank of Korea (BOK) economic indicators via the FinanceDataReader library.
[PROMPT_INJECTION]: As a data-ingestion skill, it processes external content that could theoretically harbor indirect prompt injections. However, the data consists of financial metrics and time-series observations, which pose a minimal risk of containing executable instructions for the agent context.

data-kr