Skip to main content

File connectors

Instead of downloading documents and re-uploading them, connect the place your team already keeps them — Google Drive, OneDrive, SharePoint, Dropbox, Box, an S3 bucket or an SFTP server — and import straight from there.

What you can connect​

ConnectorWhat it's forHow you connectKeeps files in sync
Google DriveImport from My Drive and Shared Drives, and save results backSign in with GoogleYes
OneDriveImport from a personal or work Microsoft accountSign in with MicrosoftYes
SharePointImport from your organisation's SharePoint document librariesSign in with MicrosoftNo
DropboxImport from your Dropbox foldersSign in with DropboxNo
BoxImport from your Box foldersSign in with BoxNo
NotionImport pages from your Notion workspaceSign in with NotionNo
JiraImport issues from your Jira projectsSign in with AtlassianNo
ConfluenceImport pages from your Confluence spacesSign in with AtlassianNo
S3 bucketImport from AWS S3 or any S3-compatible storageEnter connection detailsNo
SFTPImport from a file server you reach over SFTPEnter connection detailsNo
Google SheetsLet agents find, read, edit and create your Google Sheets, live, in chats and workflowsSign in with GoogleNot a file source
JobKred LithiumSync job profiles and company skills to a Lithium realmEnter realm and API keyNot a file source

The Connectors page groups these into File connectors — the places you import documents from — and Integrations, which is where Google Sheets and JobKred Lithium live because they aren't file sources.

To use any of them you need:

  • A paid team plan — connectors come with the paid team plans. See Plans & usage.
  • An account with the provider (or, for S3 and SFTP, the connection details below).
Connections belong to a project

You connect a source inside the project you have selected, and each project holds at most one connection per service. In a shared project, a project Admin makes the connection and every member can then import through it — the connect dialog tells you who will be able to use it before you agree. If you'd rather keep an account to yourself, connect it in your private default project instead; the dialog offers to switch you there. See Projects.

Connect a source​

  1. Open Connectors from the left sidebar.
  2. Find the card you want and choose Connect.
  3. For Google Drive, OneDrive, SharePoint, Dropbox and Box, sign in with that provider and grant access. For S3 and SFTP, fill in the details form — JobKred tests the connection before saving it.

Each card shows its state:

  • Not connected — nobody has linked this one to your account yet.
  • Connected — ready to use.
  • Reconnect needed / Expired — the provider needs you to sign in again. Choose Connect.
  • Disconnected — the connection was removed. Reconnect any time.
  • Error — the last attempt failed; try connecting again, and check that your account or keys still work.
  • Unavailable — this connector isn't switched on for your workspace yet. Ask your administrator.

Disconnecting stops new imports. Files you already imported stay in your workspace.

Storing credentials

Connection details for S3 and SFTP are encrypted and never shown again once saved. To change them, disconnect and connect again.

Import files​

Once a source is connected:

  1. Open Files and choose Add File, then pick the connector by name (the new-file form has the same options under its Integrations tab).
  2. Browse folders or search, then pick the file you want.
  3. JobKred imports it, reads it, and adds it to your Files list like any other document.
  • Source tagging — imported files are labelled with where they came from, so the Source filter on the Files page can show just those.
  • Size — imports can be up to 200 MB, which is larger than the 20 MB limit for uploading from your computer.
  • In a chat — Ellie can also list, search and pull in a file from a connected source while answering.

Keep files in sync​

Use Refresh on the Files page to re-check a connected source. JobKred reports what changed — files updated, removed at the source, or renamed — so your workspace stays current without re-uploading anything by hand.

Sync is available for Google Drive and OneDrive

The other connectors don't report changes. To pick up an edit made elsewhere, import the file again.

The first refresh after connecting records a starting point rather than a list of changes. Everything that happens from then on is detected.

Google Drive​

What you need: a Google account. Nothing to install.

Connect once by signing in with Google, then choose Google Drive when adding a file and use the Google Picker to select what you want. You can pick several files at once, from My Drive or a Shared Drive.

Least-privilege by design

JobKred only ever sees the specific files you pick — not your whole Drive. Connecting is safe and scoped.

  • Native Google formats — Docs, Sheets and Slides are converted to Word, Excel and PowerPoint automatically on import.
  • Save back to Drive — from a file's details you can export it into a Drive folder you choose. Google Drive is the only connector that supports this.
  • Sync — supported. See Keep files in sync.

Google Sheets​

What you need: a Google account. Nothing to install.

Google Sheets is a separate connector from Google Drive, and it works differently: nothing is imported. Once it's connected, the agents in your chats and in every workflow step of that project can find a spreadsheet by name, read it, change it and create new ones — live, in Google Sheets.

  1. Open Connectors, find Google Sheets and choose Connect. Sign in with Google and allow access.
  2. That's it — just ask. In a chat, or in any workflow step's instructions: "Update test_sheet with this week's numbers", "Add a row to the Hiring tracker for Ana", "Read the Q3 Budget sheet and summarise it".

There is nothing to switch on per step and no spreadsheet to pick in advance: the agent searches your Google Sheets for the name you gave, then reads and edits the one it found.

  • Finding the right sheet — if no spreadsheet matches the name, the agent tells you and asks for the exact name or link. If several match, it uses the one with exactly that name, or lists them and asks you which one.
  • What Google asks you to allow — JobKred can see the names of the files in your Google Drive (to find spreadsheets by name, never their contents) and can read, edit and create your Google Sheets.
  • Native Google Sheets only — an Excel file (.xlsx) stored in Drive isn't a Google Sheet and can't be edited this way. Open it in Google Sheets and use File → Save as Google Sheets first.
  • Changes are live — edits land in Google Sheets straight away; there is no undo from JobKred. Adding rows never overwrites what's already there.
  • Tools switches — the connection's page on Connectors has a switch for each capability: find by name, read, edit, create, and Clear and delete sheet data. The last one is off until someone who manages the connection turns it on.
  • Shared with the project — like every connection, it's used by whoever in the project runs the chat or workflow, acting as the Google account that connected it. Every action is recorded in an audit log.
Availability

Google Sheets may show as Unavailable until it is switched on for your workspace.

OneDrive​

What you need: a Microsoft account — personal, or a work or school one.

Connect by signing in with Microsoft, then browse or search your OneDrive when adding a file.

  • Work and school accounts — your IT administrator may need to approve JobKred's access before the first connection succeeds.
  • Sync — supported.
  • Disconnecting — removing the connection here clears what JobKred stored, but doesn't remove JobKred from your Microsoft account's list of connected apps. Remove it there too if you want it gone at both ends.

SharePoint​

What you need: a work or school Microsoft account. A personal Microsoft account has no SharePoint, so JobKred will ask you to sign in with an organisation account instead.

Connect by signing in with Microsoft, then browse sites → document libraries → folders, or search across every library you can open, when adding a file.

  • Administrator approval — reading SharePoint sites needs a permission most organisations only grant after an IT administrator approves JobKred's access. If connecting or browsing is refused, ask your administrator.
  • What JobKred can see — the sites and libraries you can already open with your Microsoft account; nothing more. Only the files you pick are imported.
  • Sync — not available. Imported files are point-in-time snapshots; import a file again to pick up edits made in SharePoint.
  • Disconnecting — removing the connection here clears what JobKred stored, but doesn't remove JobKred from your Microsoft account's list of connected apps.

Dropbox​

What you need: a Dropbox account.

Connect by signing in with Dropbox, then browse folders or search when adding a file.

  • Dropbox Business — you see your own member folder, not shared team spaces.
  • Sync — not available. Import the file again to pick up edits.

Box​

What you need: a Box account. A free one works.

Connect by signing in with Box, then browse folders or search when adding a file.

  • Long gaps — a connection left completely unused for around two months stops working. Reconnect from the card.
  • Sync — not available.

Notion​

What you need: a Notion account.

Connect by signing in with Notion. During sign-in, Notion asks you to choose which pages JobKred may access — JobKred only ever sees the pages you grant. Then browse those pages or search them when adding a file.

  • What imports — each page you pick becomes a readable document in your Files, ready to use in chats and workflows.
  • Sync — not available. Importing the same page again refreshes your copy rather than creating a duplicate.

Jira​

What you need: an Atlassian account with access to a Jira site.

Connect by signing in with Atlassian, then browse your Jira projects or search issues when adding a file.

  • What imports — each issue you pick becomes a readable document (summary, description, comments) in your Files.
  • Shared sign-in — Jira and Confluence use the same Atlassian sign-in but are connected separately, so connecting one doesn't connect the other.
  • Sync — not available. Importing the same issue again refreshes your copy.

Confluence​

What you need: an Atlassian account with access to a Confluence site.

Connect by signing in with Atlassian, then browse your Confluence spaces or search pages when adding a file.

  • What imports — each page you pick becomes a readable document in your Files.
  • Sync — not available. Importing the same page again refreshes your copy.

S3 bucket​

What you need: the details below. This works with AWS S3 and S3-compatible services including MinIO, Wasabi, Backblaze B2 and Cloudflare R2.

FieldRequiredDefaultWhat to enter
Endpoint URLNo—Leave blank for AWS S3. For other services, the URL the vendor gives you
RegionNous-east-1
BucketYes—The bucket to import from
Access key IDYes—A key with read access to that bucket
Secret access keyYes—
Key prefixNo—Limits browsing to keys under this folder
Use path-style addressingNotrueKeep it on for MinIO; it's harmless for AWS

JobKred checks the bucket as you save. If the key is wrong, or can't read the bucket, you'll see "Could not access bucket …" straight away.

  • Browsing — folders come from key prefixes, and you can't navigate above the prefix you set.
  • Search — filters the folder you're currently in. It isn't a search across the whole bucket.
  • Rotated keys — a key that's been rotated or disabled stops the connection working. Reconnect with the new one.
  • Sync — not available.

SFTP​

What you need: the details below, and a server JobKred can reach over the internet.

FieldRequiredDefaultWhat to enter
HostYes—Hostname or IP address
PortNo22
UsernameYes—
PasswordOne of the two—A password or a private key is required
Private key (PEM)One of the two—OpenSSH or PEM format
Private key passphraseNo—If the key is protected
Base pathNo/The folder browsing starts in

On the first connection JobKred records the server's fingerprint and checks it every time after that.

  • Browsing — starts at the base path and can't go above it.
  • Search — covers the base folder only; it doesn't look inside subfolders.
  • If the host key changes — after a server rebuild or key rotation, JobKred refuses the connection on purpose. Disconnect and reconnect to trust the new key.
  • If the server is offline — the file picker says so and offers Try again.
  • Sync — not available.

JobKred Lithium​

JobKred Lithium isn't a file source. It syncs job profiles and company skills to a Lithium realm, and can be switched on as an enrichment source in a chat.

What you need: a Professional plan or above, an administrator account, and your Lithium realm, API key and API URL.

Open Connectors, choose the Integrations category, then JobKred Lithium. Enter the three values and choose Test connection. Once it's set up, everyone in your company can toggle it on in a chat.