The Agent Skills Directory

[SAFE]: The skill operates entirely within the context of the ThinkingData platform using the ae-cli tool for all data and metadata operations, ensuring actions are performed within authorized environments.
[SAFE]: A transparent Python validation script (scripts/validate_segments.py) is used to perform deterministic integrity checks on exported data, preventing processing of malformed or inconsistent analysis results.
[SAFE]: The skill implements a rigorous Stage-1 calibration gate that requires explicit user confirmation for key business parameters such as analysis universe, time windows, and entity resolution before any data processing occurs.
[SAFE]: SQL construction instructions for the agent emphasize best practices for the target platform, such as using partition predicates ($part_date) to optimize query performance and quoting identifiers to prevent syntax errors.
[SAFE]: Strategy generation is dynamically grounded in verified project metadata, ensuring that operational recommendations are based on actual features present in the analyzed project rather than generic templates.

rfm-segmentation