db-read

Database document retrieval works by reading an index to identify candidate documents, evaluating relevance to your query, and synthesizing a concise answer with citations. If the index is empty, it reports that no documents are available.

Searching a knowledge base involves matching your query against an index of stored entries, then reading each matched document to synthesize key points. This skill reads the index, evaluates relevance, and returns a summarized answer with citations.

When the database index is empty, document retrieval reports that no documents are available. This prevents irrelevant or hallucinated results, ensuring you only receive synthesized answers when matching documents actually exist in the knowledge base.

Yes, you can locate and synthesize information from stored logs. The skill identifies relevant log entries by matching keywords or topics against the index, reads each matched document, and synthesizes a concise summary with citations.

No special formatting is required to search stored documents. You simply provide a question or query, and the skill automatically reads the index, evaluates relevance, reads matched documents, and synthesizes a concise answer with citations.

Index-based retrieval distinguishes itself by first reading an index to identify candidate documents, then reading each matched document to evaluate relevance before synthesizing a concise answer with citations, rather than scanning the entire database.