search-strategy

The search-strategy AI skill is designed to optimize enterprise search by transforming natural language questions into precise, multi-source queries. It addresses the challenge of retrieving accurate and relevant information from multiple connected sources, including chat platforms, knowledge bases, and project trackers. By breaking down complex questions into smaller, targeted searches, it ensures that results are both comprehensive and contextually appropriate. This skill is particularly valuable in environments where information is scattered across various systems and formats, reducing the manual effort needed to locate specific answers.

The skill offers several key capabilities. It classifies queries by type—such as decisions, status updates, documents, personnel, factual information, temporal events, and exploratory questions—to determine the most effective search strategy. It extracts critical search components like keywords, entities, intent signals, constraints, and negations from the user query, enabling precise targeting of results. Additionally, it generates sub-queries tailored for each source, choosing between semantic or keyword-based searches based on query characteristics, and produces multiple query variants to account for different terminology or phrasing. Finally, it synthesizes results from all sources into a coherent, ranked, and deduplicated output.

Use cases for this skill include locating decision outcomes, tracking project statuses, finding specific documents, identifying responsible personnel, confirming organizational policies, and conducting exploratory research across enterprise systems. Target users are knowledge workers, project managers, engineers, and analysts who require efficient, accurate access to information spread across multiple platforms. By streamlining query decomposition and multi-source search orchestration, this skill enhances productivity and ensures informed decision-making.