Intelligent Extraction & Ingestion FAQs

Question 1

What types of archival sources can it process?

Accepted Answer

It is designed to handle email archives, correspondence, receipts, and historical documents, applying selective logic to extract significant events while skipping administrative noise.

Question 2

How does the extract skill handle sensitive information?

Accepted Answer

The skill follows a strict privacy protocol, routing sensitive raw data to a secure private directory ($ACA_DATA) and ensuring only depersonalized, generalized principles are added to public-facing frameworks.

Question 3

Does it prevent duplicate entries in my knowledge base?

Accepted Answer

Absolutely. It performs a mandatory search of your Personal Knowledge Base (PKB) before creating new content, ensuring that existing information is augmented rather than duplicated.

Question 4

Can I extract training data from tracked changes in Word documents?

Accepted Answer

Yes, the skill includes specialized logic for processing DOCX files with tracked changes, converting them to preserve markup and extracting feedback-source pairs for model training.

Question 5

What is the difference between /extract and specialized skills like /remember?

Accepted Answer

The /extract skill acts as an intelligent router. Use it when you have mixed inputs or are unsure of the extraction goal; it will analyze the data and delegate to specialized skills like /remember or /decision-extract automatically.

Intelligent Extraction & Ingestion

主要功能

使用场景

Intelligent Extraction & Ingestion

主要功能

使用场景