The Agent Skills Directory

Proactive Security Guidance: The skill includes a specific warning to treat external extracts as untrusted data. This is a recommended practice to prevent indirect prompt injection when processing third-party information.
Data Normalization and Sanitization: The instructions emphasize coercing data types and aligning formats during the normalization step. This process helps ensure data integrity and prevents issues related to malformed input processing.
Absence of Risky Operations: The skill is purely instructional and does not perform network requests, file system modifications, or execution of external code.

gl-recon