The Agent Skills Directory

[SAFE]: The skill implements a legitimate workflow for financial quantitative research. The provided Python script (scripts/transaction_cost_analysis_global_diagnostics.py) performs deterministic statistical analysis on local CSV files using the standard Pandas library. It does not perform any network operations, access sensitive system paths, or utilize dynamic execution patterns.
[INDIRECT_PROMPT_INJECTION]: The skill includes a surface for processing untrusted data through the ingestion of external CSV files.
Ingestion points: scripts/transaction_cost_analysis_global_diagnostics.py reads data from a user-provided input.csv via pd.read_csv.
Boundary markers: None explicitly defined in the prompts, though the script enforces a strict column schema.
Capability inventory: The skill is limited to local file system read/write operations for the input data and the resulting JSON diagnostics.
Sanitization: The Python script utilizes pd.to_numeric(errors="coerce") and pd.to_datetime(errors="coerce") on all critical fields. This effectively sanitizes the input by converting any potential malicious prompt injection strings into NaN values before they are used in calculations or written to the output payload.

transaction-cost-analysis-global