Files attached to records are one use case, and a common one: a contract attached to a customer, a proof of delivery attached to an order, a resume attached to an applicant. The file lives with the record; the record is where users go to find it. But there's a second use case that doesn't fit that shape at all. A shared library of documents that belong to the tenant rather than to any particular record. Corporate policies, legal templates, marketing assets, reference material, standard operating procedures, onboarding materials, image libraries. These aren't attached to anything specific. They belong in folders, they get organized by topic, they get shared by link, they get permissioned at the folder level. The file repository is purpose-built for that use case.
Structurally, the repository is familiar: a hierarchy of folders and sub-folders, with files as the objects within them. Users navigate with breadcrumbs and a folder tree; they drag-to-move files and folders around; they see recent items for quick access to files they've been working with. The experience is what a decent file manager feels like, adapted to the web and integrated with the rest of the platform.
Each file in the repository is a managed object — a tracked entity with metadata, permissions, version history, and a place in the activity log. That's the critical difference from storing files as ad-hoc blobs in the cloud bucket of your choice: the file repository isn't just storage, it's a first-class part of the platform with all the platform's machinery around it. A file in the repository can be referenced from records, attached to emails automatically, processed by automations, exposed through the API, and audited the same way any other entity is audited.
Access control is layered: permissions at the folder level cascade to the files inside, and per-file permissions can adjust the default for specific items. A folder for HR documents can be visible only to users in the HR role, and anything placed inside that folder inherits the restriction automatically. A particular file within that folder can be made accessible to additional roles if it needs broader visibility. Moving a file between folders re-evaluates its effective permissions, so organizing by folder is also organizing by access — which is usually exactly what users are already thinking about when they put a file in a folder named "Legal."
Bulk operations handle the common cases where one file at a time would be tedious. Select a set of files; move, copy, or delete them as a group. Rename a folder and the files inside carry their context forward. Import a batch of files into a folder from a single upload action. These are conveniences that matter at any real scale — a repository with a hundred files is manageable one click at a time; a repository with ten thousand files needs bulk.
Search within the repository is full-text over file names, metadata, and — where the file format allows — file contents. Find a document by a phrase inside it, not just by its filename. The same indexing machinery that powers search across the rest of the platform powers the repository, which means permissions are respected (users don't see files they can't access, even as search results) and results are fast at scale.
The media library and the repository coexist. The media library holds files that are attached to records, cataloged for reuse across record-bound contexts — product images referenced by multiple orders, company logos reused across email templates, marketing collateral embedded in canvas pages. The repository holds standalone files that belong to the tenant rather than to any particular record. The two libraries are distinct because they serve different workflows; files can also be promoted from one to the other when a use case evolves.
Programmatic access from automations turns the repository into part of the workflow fabric rather than a purely human-facing module. A nightly automation can export a report and drop it into a Monthly Reports folder. A workflow can pick up a file that was placed in an Incoming folder, process it, and move it to Processed. A scheduled task can generate fresh versions of a templated document and replace the existing version. The repository is a target and source for automation, the same way records are.
For code that needs deeper control, the platform's internal API exposes repository operations to server-side scripts: read files, write files, create folders, manage metadata. That's how more sophisticated integrations — custom importers, bridges to external document systems, scheduled maintenance — interact with the repository. The same operations are available to external systems through the REST API, so partner integrations that want to push documents into the repository or pull from it can do so with the same idioms as any other resource.
Versioning and auditing integrate with the platform's broader mechanisms. A file that's updated preserves its previous version; users can see what changed, when, and by whom. The activity log records every operation — creates, updates, moves, deletes, permission changes — so the forensic question "who replaced the Q3 contract last Tuesday?" has a one-click answer. The version-management article covers the platform's broader approach to controlled change; the repository participates in it naturally.
Default view modes — list and grid — let users pick the presentation that fits their working style, per user. The resizable sidebar keeps the folder tree at whatever width works for the user's screen. These are small conveniences in isolation; in the aggregate, they're what turn a functional file system into one people actually enjoy using.
For tenants with substantial document-management needs — legal firms, professional services, HR-heavy organizations, any team that trades documents throughout the day — the file repository is often the feature that consolidates several disparate tools into one. Instead of a corporate cloud drive, a separate document management system, and a file server that's been "temporary" for eight years, the tenant has a repository inside the platform where the files live alongside the records, the workflows, and the people who use them. That consolidation is worth more than any individual feature in isolation; it's what makes the platform the single place work actually gets done.