The Agent Skills Directory

[SAFE]: The skill provides structured instructions for product analysis that prioritize data integrity and privacy. It includes explicit guidelines to minimize data selection and avoid exposure of sensitive information like PII.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes external datasets such as event logs and qualitative feedback. This is a core functionality and is governed by mandatory data-quality audits and privacy-preserving instructions. Ingestion points: Event logs, schema metadata, and qualitative evidence. Boundary markers: Structural separation between analysis results and query plans. Capability inventory: Generation and execution of SQL queries. Sanitization: Instructions for data aggregation and pseudonymization in the Privacy and safety section.
[DYNAMIC_EXECUTION]: The skill generates SQL queries at runtime using established templates. These templates utilize parameter placeholders which support secure query generation.

user-insights