The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill processes external text data (patent claims), creating a surface for potential indirect prompt injection attacks. • Ingestion points: The Read tool is used to ingest user-provided text files. • Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the logic. • Capability inventory: The skill utilizes Bash, Read, and Write tools. • Sanitization: No input validation or sanitization of the claim text is performed before it is passed to the analyzer.
[Command Execution] (SAFE): The Bash tool is used to execute a local script (claims_analyzer.py) consistent with the skill's stated purpose.
[Data Exposure & Exfiltration] (SAFE): No access to sensitive system directories or credentials was found, and no external network calls are initiated.

patent-claims-analyzer