The Agent Skills Directory

[SAFE]: The skill provides standard Python logic for legal document templating and compliance checks without any signs of malicious code, data exfiltration, or obfuscation.
[INDIRECT_PROMPT_INJECTION]: The skill accepts untrusted user input (such as party names and service descriptions) to generate legal documents, which presents a surface for indirect prompt injection.
Ingestion points: Data enters the skill through parameters in generate_nda, generate_service_agreement, and the assess_gdpr_compliance dictionary in SKILL.md.
Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the document templates.
Capability inventory: The skill logic is restricted to string formatting and dictionary processing; no dangerous functions like eval, exec, or network operations are used in the provided scripts.
Sanitization: The skill does not perform sanitization or validation on inputs before interpolating them into document templates.

lawyer-expert