search-strategy

Multi-source search decomposes a natural language query into parallel, source-specific sub-queries, executing them across connected data stores and synthesizing ranked, deduplicated results into a single coherent answer.

Ambiguity handling detects uncertainty in cross-source queries and applies fallback strategies to ensure useful answers. It orchestrates source-specific translation and ranking across documents and trackers to resolve conflicting or vague information.

Query decomposition splits a natural language question into targeted, per-source sub-queries. It translates each sub-query into the specific syntax required by chat logs, knowledge bases, and trackers before running parallel searches.

Yes, source translation automatically converts a single user question into the correct syntax for each connected data store. It executes parallel searches across documents and trackers, then deduplicates and ranks the combined results.

Multi-source search applies deduplication and relevance ranking to synthesized results from parallel searches. If duplicates persist, the ambiguity handling and fallback strategies may need adjustment to better resolve conflicting source data.

Yes, connected data sources are required. The skill orchestrates parallel, source-specific searches across chat logs, knowledge bases, project trackers, and documents by translating a single query into the appropriate syntax for each connected store.