This skill orchestrates a safe pipeline to extract structured tabular data from non-CSV sources (images, PDF, Word, Excel), map fields against a knowledge network (ontology), produce a validated CSV, and import rows via the platform's data-flow import endpoint. It enforces strict safety rules: no direct DB connections, only INSERT via ontology ds import-csv, and explicit user confirmation before writes.
Use when you need to convert scanned forms, documents, or spreadsheets into structured rows and load them into the platform's data tables while preserving field semantics and platform auditing. Triggers include phrases like '数据采集', '从文档提取数据', '图片转数据', '数据导入'.
Designed for CLI-based ontology tooling and agent runtimes that can call external skills and shell helpers (ontology CLI / ontology-core).
This skill has not been reviewed by our automated audit pipeline yet.