The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill utilizes an on-the-fly Python script generation approach to handle bespoke financial logic for individual scenarios. Safety is addressed by a mandatory testing policy requiring zero-volatility and conservation checks for all generated code before it is trusted.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, such as bank statement exports, CSV files, and user-provided notes. This creates an attack surface where malicious instructions could be embedded in data to influence the agent's analysis or modeling outputs.
[COMMAND_EXECUTION]: The instructions direct the agent to execute shell commands to run unit tests and verify the integrity of the modeling environment.

financial-planning