The Agent Skills Directory

[SAFE]: The skill is primarily composed of high-quality markdown documentation regarding fintech compliance, KYC/AML, and embedded finance. The provided script, scripts/regulatory_trigger_checker.py, was analyzed and found to be a safe utility that uses standard Python libraries (argparse, json, re) to perform regex-based keyword matching on local text files. It contains no network operations, subprocess executions, or dynamic code evaluation.
[PROMPT_INJECTION]: The skill includes an analysis script that processes external text files provided by the user, which represents a surface for indirect prompt injection. However, the script has no dangerous capabilities, and the risk is assessed as safe.
Ingestion points: scripts/regulatory_trigger_checker.py reads the content of a business description file provided via the command line.
Boundary markers: Absent; the script output does not use specific delimiters to isolate user-provided data.
Capability inventory: No network access, file system modifications, or shell execution capabilities were found in the skill's code or scripts.
Sanitization: Absent; the script processes input files as raw text for regex comparison.

fintech-advisor