The Agent Skills Directory

Internal Script Execution: The skill facilitates its workflow by executing local Python scripts (user_context_preflight.py and validate_payload.py) for state management and dashboard validation. These operations are performed within the plugin's internal directory structure and represent standard functional logic.
Data Ingestion and Indirect Prompt Injection: The skill is designed to ingest and process data from various external sources, including calendar invites, emails, chat threads, and financial filings. While this creates a potential surface for indirect prompt injection, the skill mitigates this risk through a comprehensive source hierarchy, mandatory citation policies, and explicit labeling of verified facts versus management claims.
Compliance and Safety Controls: Extensive instructions are provided to ensure the agent adheres to financial compliance standards, specifically prohibiting the solicitation of Material Non-Public Information (MNPI) and maintaining a clear distinction between internal strategy and external-facing materials.

meeting-prep