The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides shell commands for the agent to manage its environment and execute data retrieval tasks. This includes activating a virtual environment, installing dependencies from a local requirements file, and running Python scripts from a shared directory. Evidence found in references/data-queries.md.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze financial market data from external providers, which constitutes an attack surface for indirect prompt injection.
Ingestion points: Market data is retrieved from external sources such as EastMoney and AKShare through the views_runner.py script as described in SKILL.md and references/data-queries.md.
Boundary markers: No specific boundary markers or delimiters were identified in the instructions for isolating external data from the agent's core instructions.
Capability inventory: The skill utilizes shell command execution and Python script execution to perform its primary data analysis functions.
Sanitization: The provided instructions do not specify any sanitization, validation, or filtering of the external data before it is processed by the agent.

northbound-flow-analyzer