AI Connection Security

Giving an AI agent access to your data deserves a clear security model. Here's ours.

Tokens

• Created by you, in the dashboard. Each connection uses a personal access token you generate and name.
• Shown once. The token value is displayed a single time at creation; only a hash is stored. If you lose it, you revoke it and make a new one.
• Workspace-scoped. A token can only ever reach the workspace it was created in — never another workspace, and never anyone else's data.
• Capped and revocable. You can keep several active tokens and revoke any of them instantly from the dashboard. Tokens stop working immediately if the user who created them leaves the workspace.
• Rate-limited against runaway clients.

Access levels

Every token is created with one of two levels:

• Read-only — the agent can analyze tables, query data, and check run status. It cannot change anything or spend anything.
• Full access — adds cleanup operations, starting/stopping runs, and creating new extractions.

Start read-only; upgrade to full access when you trust the workflow.

Credit protection

• Reading, querying, cleaning, and exporting data consume no credits.
• Anything that would spend credits (starting an extraction) returns a cost estimate first so the agent can confirm with you, and runs are refused when the balance is empty — balances can't go negative.

Privacy boundaries

• Local data stays local. Agents can only see cloud data. Tables in the Chrome extension's local storage are never exposed.
• No cross-workspace access. Team members' other workspaces are invisible to a token.
• Large-response protection. The server refuses to dump oversized data into a conversation and asks the agent to narrow its query instead.

Good practices

1. Create one token per AI client, named after it ("Claude Desktop", "Cursor — laptop"), so revocation is surgical.
2. Prefer read-only tokens for analysis-only workflows.
3. Revoke tokens you no longer use — creating a fresh one takes seconds.