Provider REST API

17 actions

Connect Claude to Neon

Manage serverless Postgres projects, branches, databases, roles and computes. Toolspoke puts 17 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: 17
What a call bills: Nothing
Adapter: Maintained by Toolspoke

Connected in three steps

1
Install Neon
Open the marketplace in your workspace, add Neon 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. Neon 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 Neon 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: Neon Console → your avatar → Account settings → API keys → Create new API key for a personal key, or Organization settings → API keys for a key that reaches only that organization's projects.; napi_…

What Claude can do in Neon

17 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: 10Reads
Writes: 5Writes
Destructive: 2Destructive

Reads

Fetches data and changes nothing.

list_projects
List the Neon projects this key can reach, with their region, Postgres version and default branch. Takes no required arguments, which is why an install tests itself with it. A personal key that belongs to several organizations may need org_id to see an organization's projects.
get_project
Read one project's settings, region, Postgres version, compute defaults and storage and compute usage for the current period.
list_organizations
List the organizations the key's owner belongs to, with the org_id each project listing can be narrowed by.
list_branches
List a project's branches with their parent, state, size and whether each is the default. A branch id (br-…) is what every database, role and restore call takes.
get_branch
Read one branch, including its parent, the point it was branched from, its logical size and its protection state.
list_databases
List the Postgres databases on a branch with the role that owns each.
list_roles
List the Postgres roles on a branch. Passwords are not returned here.
get_connection_uri
Return a ready-to-use postgresql:// connection string for a database and role. The string contains the role's password, so this response is never written to the audit log.
list_endpoints
List a project's computes (endpoints) with their branch, host, state (active or idle), size limits and autosuspend setting.
list_operations
List a project's recent control-plane operations (branch creation, compute starts, restores) with their status. Use it to see whether a change has finished or why it failed.

Writes

Creates or updates something on the other side.

create_branch
Create a branch: an instant copy-on-write copy of its parent's data, so it is cheap and can be deleted afterwards. Forks the project's default branch unless parent_id is given. Adds a read-write compute by default so the branch can be connected to at once; set compute_type to read_only for a replica compute instead.
create_database
Create an empty Postgres database on a branch, owned by an existing role. Reversible by deleting it in the console; nothing existing is changed.
start_endpoint
Start an idle compute. Computes also wake on their own when a connection arrives, so this is for warming one up ahead of traffic.
suspend_endpoint
Suspend a compute now. Open connections are closed; the next connection starts it again.
restart_endpoint
Restart a compute, for example to apply new settings. Open connections are dropped while it restarts.

Destructive

Deletes or permanently alters something. Worth granting on purpose.

delete_branch
Delete a branch with its databases, roles and computes. Destructive because the branch's data goes with it and a branch that has children cannot be deleted at all. The default branch cannot be deleted.
restore_branch
Restore a branch's data from a source branch, optionally at a point in time. Destructive because it replaces the target branch's current data: set preserve_under_name to keep the old state as a backup branch, otherwise it is gone. Restoring a branch to its own earlier state needs source_lsn or source_timestamp.

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 17 actions above are the whole of it. A call to any other name is refused before it reaches Neon 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 Neon 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 Neon: 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 Neon. An agent has to ask.
It does not smooth over provider limits: Toolspoke does not retry, queue or back off around Neon's own rate limits. A call that Neon refuses comes back to the agent as a failed call.

Before you connect it

What can Claude do in Neon?

17 named actions: 10 that only read, 5 that write and 2 that delete or permanently alter something. They include list_projects, get_project and list_organizations. 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 Neon 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 Neon connector work with Cursor and Codex, or only Claude?

Any client that speaks MCP, and every one of them gets the same 17 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 Neon connector not do?

The 17 actions above are the whole of it. A call to any other name is refused before it reaches Neon 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 Neon 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 Neon. An agent has to ask. Toolspoke does not retry, queue or back off around Neon's own rate limits. A call that Neon 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 Neon'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 Neon?

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 database connectors

Browse all of them