The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches financial market data from well-known and reputable services including Tencent (gtimg.cn), Sina (sina.com.cn), and Eastmoney (via the akshare library). These operations are standard for the skill's purpose of stock analysis.
[COMMAND_EXECUTION]: The skill uses Python's ThreadPoolExecutor in scripts/daily_decisions.py and scripts/realtime_quotes.py to perform concurrent network requests for stock data, which is an efficient implementation for scanning large stock universes.
[DATA_EXPOSURE]: The script scripts/daily_decisions.py allows reading a local file via the --holdings argument to check for sell signals. It specifically extracts 6-digit stock codes and does not attempt to access sensitive system files or credentials.
[INDIRECT_PROMPT_INJECTION]: The skill has a potential surface for indirect injection as it processes data from external APIs.
Ingestion points: MarketDataProvider.get_quote, get_history, and get_mainboard_universe in scripts/paper_trading/market_data.py fetch data from gtimg.cn, sina.com.cn, and akshare.
Boundary markers: Absent; the skill does not explicitly instruct the agent to ignore instructions embedded in stock names or metadata.
Capability inventory: Uses ThreadPoolExecutor for parallel processing, requests.get for network access, and Path.read_text for local file ingestion.
Sanitization: The skill validates and coerces numeric fields (prices, volumes), but passes stock names as strings directly into the output payload without specific escaping for LLM context.

a-share-strategy-mainboard-multi-swing-defensive