search

The search skill provides a unified way to search across multiple connected MCP sources using a single query. It is designed to help users quickly locate decisions, documents, conversations, status updates, and other information that may be distributed across chat platforms, email systems, cloud storage, project trackers, CRM tools, or knowledge bases. Instead of manually searching each service independently, the skill analyzes the user’s query, identifies intent and filters, and coordinates searches across all available connected sources. This reduces the time and effort required to find important organizational knowledge and historical context.

The skill supports query parsing, source-aware search translation, parallel execution, and result synthesis. It can extract entities, time constraints, source hints, and explicit filters such as from:, in:, after:, before:, and type:. For each connected source, it generates targeted sub-queries using the source’s native search capabilities. It also supports semantic and keyword-based search approaches depending on the platform. Results from different systems are collected simultaneously, ranked for relevance, and deduplicated to avoid repeating the same information across multiple sources. The workflow is designed to continue operating even if one or more connected sources fail or return errors.

This skill is useful for teams and professionals who work across multiple communication and productivity platforms and need a centralized way to retrieve information. Common use cases include finding past project decisions, locating documents mentioned in discussions, searching for emails related to a topic, identifying conversations from a specific timeframe, or tracing updates across project management tools. It is particularly valuable for organizations with distributed knowledge spread across collaboration tools, storage systems, and operational platforms.