# Tokens and security

> How access tokens work, and keeping them safe.

Everything an agent does in Flisk, it does as one person, with an access token of theirs, on their screens only.

## Making a token

Make one token for each agent, named for what will use it:

- in the Flisk app: Settings, Connect your agent;
- or at a terminal: `flisk tokens new "Claude, laptop"` (copied to the clipboard, never printed, and shown once).

Accounts are made in the Flisk app; an agent can't make one.

## Using it

Send it with every request to `https://flisk.tv/mcp`:

```
Authorization: Bearer <token>
```

An agent on the person's own computer needn't have one: `flisk connect` runs Flisk's MCP server through the command
line, signed in as it is.

## Keeping it safe

- A token acts as the person, on their screens only. It can't reach anyone else's screens, or read their messages
  to other people.
- Treat it like a password: never in a public repository, a shared file or a log.
- `flisk tokens` lists them; `flisk tokens rm ID` revokes one, and anything using it stops at once.
