The Agent Skills Directory

[COMMAND_EXECUTION]: The skill directs the agent to execute shell commands to activate a virtual environment and run data processing scripts located in the sibling directory ../findata-toolkit-cn/scripts/. These operations are required for the skill's primary analytical functions.
[EXTERNAL_DOWNLOADS]: The skill instructions involve installing Python dependencies via pip and fetching financial data from established services such as AKShare and Eastern Wealth. These resources are from well-known and trusted technology/financial providers.
[PROMPT_INJECTION]: The skill handles untrusted external financial data, which presents an indirect prompt injection surface. This is mitigated by defined analysis workflows and structured response constraints.
Ingestion points: Financial data is ingested from AKShare and Eastern Wealth APIs as detailed in references/methodology.md.
Boundary markers: The agent is instructed to follow a specific structure defined in references/output-template.md to limit response variability.
Capability inventory: The skill utilizes local Python script execution for data retrieval and processing as outlined in references/data-queries.md.
Sanitization: The methodology includes specific protocols in references/methodology.md for detecting and handling missing or abnormal data values.

market-breadth-monitor