Skip to main content

Overview

The Knowledge Repository acts as the centralized storage backbone of the platform. It provides a structured environment for managing the data that fuels your agents specifically File Search Agents. With a built in file system and document editor, it serves as the primary source of truth for your organizational intelligence.

Storage & Data

The repository is built on robust storage engines that allow for flexible data organization and retrieval.

Files and Folders

You can maintain a clean hierarchy of data by creating folders and sub-folders. This organization is critical when mapping specific directories to agents for targeted information retrieval.
  • Direct Uploads: Support for various file formats including PDF, TXT, and CSV.
  • Folder Mapping: Group related documents (e.g., “Legal Contracts 2026”) to limit an agent’s search scope.

Native Document Creation

Beyond just uploading external files, you can create and edit documents (type .docx) directly within the platform.
  • In-Platform Editing: Draft, refine, and save documentation without leaving the interface.
  • Real-time Synchronization: Changes made to native documents are immediately available for agents to index and search.

Security and Privacy

Data integrity and confidentiality are prioritized through rigorous security protocols.
  • Encryption at Rest: All files stored within the Knowledge Repository are encrypted using industry-standard protocols.
  • Access Control: Files are restricted to intended users only. Permissions ensure that sensitive data is not accessible to unauthorized team members or unauthorized agents.
  • User-Level Permissions: Define who can view, edit, or delete specific documents or entire folder structures.

File Operations

The platform provides a suite of management tools for every stored asset:
OperationDescription
UploadBring external intelligence into the platform via drag-and-drop.
DownloadExport files or native docs for external use or offline review.
DeletePermanently remove outdated or redundant data from the repository.
RenameUpdate file naming conventions for better organization.
MoveRelocate files between folders to maintain repository structure.

Agent Integration

The Knowledge Repository is the primary data source for the File Search Agent.
  1. Select Source: When configuring a File Search Agent, you point it toward a specific folder in the Knowledge Repository.
  2. Contextual Retrieval: The agent uses the encrypted files in that folder to answer queries based on real, uploaded data.

Deleting a folder or file that is currently being used by an active agent may cause workflow failures. Always check agent dependencies before removing critical data.