Provider REST API

13 actions

Connect Claude to Browserbase

Run headless browsers in the cloud, and fetch or search the web through Browserbase. Toolspoke puts 13 of its actions behind one MCP endpoint that Claude, Cursor, and Codex all speak.

Connect View pricing

Connection: Provider REST API
Authentication: API key
Actions exposed: 13
What a call bills: Nothing
Adapter: Maintained by Toolspoke

Connected in three steps

1
Install Browserbase
Open the marketplace in your workspace, add Browserbase to the project your agents work in, and it appears on the gateway immediately.
2
Connect the credential
Authenticate with api key. Where to get one, and what it has to be able to reach, is the next section.
3
Point your agent at the gateway
Give your client one address, https://toolspoke.com/mcp. Claude Code takes it as a command, Claude and Claude Desktop add it as a custom connector, and Cursor, Codex and VS Code each read it from a config file of their own.

.mcp.json

{
  "mcpServers": {
    "toolspoke": {
      "type": "http",
      "url": "https://toolspoke.com/mcp"
    }
  }
}

One block covers every tool you have installed. Browserbase shows up in the client as soon as your policy allows it, and so does everything else you install later.

Where the address goes, per client

Claude Code: Run it in your project, then /mcp to sign in
claude mcp add --transport http toolspoke https://toolspoke.com/mcp
Claude and Claude Desktop: Settings, then Connectors, then Add custom connector
https://toolspoke.com/mcp
Cursor: ~/.cursor/mcp.json, or .cursor/mcp.json for one project
{ "mcpServers": { "toolspoke": { "url": "https://toolspoke.com/mcp" } } }
Codex: ~/.codex/config.toml
[mcp_servers.toolspoke] url = "https://toolspoke.com/mcp"
VS Code: .vscode/mcp.json, or the MCP: Add Server command
{ "servers": { "toolspoke": { "type": "http", "url": "https://toolspoke.com/mcp" } } }

What Browserbase asks for

API key. You provide it once, when you install the connector. Toolspoke encrypts it at rest and decrypts it only for the length of a single call, and the gateway attaches it to the outbound request itself, so it is never part of the arguments an agent sends.

API keyRequired: browserbase.com → sign in → Overview (the dashboard home page) → API Key. Copy the key shown there.; bb_live_…

What Claude can do in Browserbase

13 actions, each one declared and named by the connector rather than discovered at runtime. A workspace policy grants a person all of them, a hand-picked selection, everything on the read side, everything on the write side, or none.

Reads: 9Reads
Writes: 2Writes
Destructive: 2Destructive

Reads

Fetches data and changes nothing.

list_projects
List the projects this key can reach, with id, name, default session timeout and concurrency. The project id is what get_project_usage takes.
get_project_usage
Read a project's usage so far this billing period - browser minutes and proxy bytes. Use it to check spend before starting many sessions.
list_sessions
List browser sessions, newest first, with id, status, region, start and end time, and proxy bytes used. Filter by `status` (RUNNING to see what is live and billing right now), or search user metadata with `q`. This is the connection test: it reads the account's own sessions, not just the key.
get_session
Read one session - status, region, timings, keepAlive, context and user metadata.
get_session_live_urls
Get the live-view URLs of a running session, for watching or taking over the browser by hand. Anyone with these URLs can control the browser, so the payload is kept out of the audit log. Set `expiresIn` to make them expire sooner than the session does.
get_session_logs
Read a session's browser logs - the DevTools Protocol requests and responses. Use it to find out why an automation failed. The logs carry the pages the browser visited, so the payload is kept out of the audit log.
get_context
Read one context's id, name, project and dates.
fetch_page
Fetch one URL through Browserbase's network, with a real browser user agent, and return its status, headers and content - as-is, as markdown, or as JSON in the shape of `schema`. No browser is started and no JavaScript runs, so a page that renders client-side comes back empty; use the MCP route for those. Content over 5 MB is refused. The content is other people's, so the payload is kept out of the audit log.
web_search
Search the web and get back ranked results with title, URL and snippet. Use it to find the page before fetching or browsing it. The results are other people's pages, so the payload is kept out of the audit log; the query is kept out too.

Writes

Creates or updates something on the other side.

create_session
Start a cloud browser and return its id and connectUrl. Nothing here drives it: hand the connectUrl to a Playwright, Puppeteer or Stagehand client, or use the MCP route instead. The browser bills by the minute until it is released or times out, so set a short `timeout` and call release_session when done. Classified write because it starts a billed resource.
create_context
Create a context - a saved browser profile that keeps cookies and logins between sessions. Pass its id in create_session's browserSettings.context to use it.

Destructive

Deletes or permanently alters something. Worth granting on purpose.

release_session
End a running session now instead of waiting for its timeout. Classified destructive because it closes the browser under whatever client is using it, and that client's work in progress is lost. Use it on sessions you started; check list_sessions first.
delete_context
Delete a context and the cookies and logins saved in it. It cannot be recovered, and sessions that relied on it start signed out.

What it will not do

Enforced by the gateway rather than left to convention, which is why each of these can be stated flatly.

It cannot call anything else: The 13 actions above are the whole of it. A call to any other name is refused before it reaches Browserbase rather than forwarded on, and connecting your account does not add to the list: it is fixed by the connector, not discovered at run time.
It reaches no further than your credential: Toolspoke holds no access to Browserbase of its own. Every call carries the credential you stored and nothing besides, so whatever that credential cannot reach, this connector cannot reach either.
It never hears from Browserbase: Nothing is pushed to it. There is no webhook, no subscription and no polling, so this connector cannot notice by itself that something changed in Browserbase. An agent has to ask.
It does not smooth over provider limits: Toolspoke does not retry, queue or back off around Browserbase's own rate limits. A call that Browserbase refuses comes back to the agent as a failed call.

Before you connect it

What can Claude do in Browserbase?

13 named actions: 9 that only read, 2 that write and 2 that delete or permanently alter something. They include list_projects, get_project_usage and list_sessions. Nothing outside that list is reachable: the connector declares each operation by name rather than proxying whatever an agent asks for.

What credentials does the Browserbase connector need?

API key. The connector asks for api key. Values are encrypted at rest and attached to the outbound request by the gateway, so they are never part of the arguments an agent sends and never reach the audit log.

Does the Browserbase connector work with Cursor and Codex, or only Claude?

Any client that speaks MCP, and every one of them gets the same 13 actions. There is a single address, https://toolspoke.com/mcp. Claude Code adds it with claude mcp add --transport http, Claude and Claude Desktop take it as a custom connector in settings, Cursor reads it from .cursor/mcp.json, Codex from ~/.codex/config.toml, and VS Code from .vscode/mcp.json. Each of them signs in to the gateway itself, so there is no key to paste.

What does the Browserbase connector not do?

The 13 actions above are the whole of it. A call to any other name is refused before it reaches Browserbase rather than forwarded on, and connecting your account does not add to the list: it is fixed by the connector, not discovered at run time. Toolspoke holds no access to Browserbase of its own. Every call carries the credential you stored and nothing besides, so whatever that credential cannot reach, this connector cannot reach either. Nothing is pushed to it. There is no webhook, no subscription and no polling, so this connector cannot notice by itself that something changed in Browserbase. An agent has to ask. Toolspoke does not retry, queue or back off around Browserbase's own rate limits. A call that Browserbase refuses comes back to the agent as a failed call.

Can I limit which actions an agent can call?

Yes, in two places. The project switches Browserbase's actions on and off one at a time, for everyone in the project at once, and the screen groups them by read, write and destructive so turning off everything that deletes is one click. An individual agent key can then be narrowed further, to particular toolkits in a project and to particular actions in a toolkit. Whatever it was granted, a key never reaches a project its owner cannot.

What gets recorded when an agent calls Browserbase?

Every attempt, with the agent that made it and the person that agent belongs to, the full request payload, the response payload, the status, the duration, and what the call cost in sandbox time and model tokens. Values whose key names a secret are masked out before the record is shown to anyone. An operation the connector marks as not retained never has its response body written at all, so the gateway keeps no second copy of what was read.

More developer tools connectors

Browse all of them