Mercury for AI agents

Mercury is business banking software, used here only for read-only lookups of accounts, balances, transactions, cards, recipients, and statements, with no ability to move money. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.

Finance and accountingOAuth 2.0 (MCP)MCP + RESTmercury.com
AUDIT LOG · MERCURYPOLICY: acme-support
09:41:07 · claude · u_8f2read
mercury.list_credit
Listcredit✓ allowed · 212ms
09:41:08 · claude · u_8f2read
mercury.get_user
Getuser✓ allowed · 164ms
EVERY MERCURY CALL, ON THE RECORD
01 · USE CASES

What agents do in Mercury.

01

Check an account's balance

Look up a business account's current balance before answering a cash-flow question.

02

Pull recent transactions

List recent transactions on an account to reconcile spend, a read-only view since Mercury blocks all writes here.

03

Review a pending money-movement request

Check the status of a pending send-money approval request without approving or changing it.

02 · ACTIONS

30 Mercury actions, graded by risk.

Every Mercury action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.

read

30

Look things up. Allowed by default.

  • mercury.get_user
    Getuser
  • mercury.get_users
    Getusers
  • mercury.get_account
    Getaccount
  • mercury.get_invoice
    Getinvoice
  • mercury.get_webhook
    Getwebhook
  • mercury.list_credit
    Listcredit
  • mercury.get_accounts
    Getaccounts
  • mercury.get_customer
    Getcustomer
  • mercury.get_treasury
    Gettreasury
  • mercury.get_webhooks
    Getwebhooks
  • mercury.get_recipient
    Getrecipient
  • mercury.list_invoices
    Listinvoices
  • mercury.get_attachment
    Getattachment
  • mercury.get_recipients
    Getrecipients
  • mercury.list_customers
    Listcustomers
  • mercury.get_transaction
    Gettransaction
+ 14 MORE

write

0

Create and change things. Allow, or ask the user first.

  • No write actions.

destructive

0

Delete, cancel or archive. Ask first, or deny outright.

  • No destructive actions.
03 · HOW IT WORKS

Mercury in three steps.

  1. 01Your users connect MercuryThey sign in to Mercury on Arc0 Connect, under your brand, and approve the access you ask for.
  2. 02You set the rulesReads run, and writes like “updates” can wait for the user to approve.
  3. 03Any agent can actYour agent calls Mercury through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
POLICY.TS
await arc0.policies.set('mercury', {
  read: 'allow',
  write: 'ask',        
  destructive: 'deny',  
})

# Claude Code: the same connection, one URL
$ claude mcp add --transport http arc0 \
    https://mcp.arc0.ai/u/u_8f2
04 · AUTH AND DATA

How Mercury connects.

Mercury runs its own MCP server behind OAuth. Arc0 registers the client, users approve access on Arc0 Connect, and your agent reaches Mercury through the same endpoint, policies and audit log as every other app.

The same Mercury connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →

AUTH
OAuth 2.0 (MCP)
CREDENTIALS
Per-tenant encrypted vault
MODEL SEES
Results only, never credentials
AUDIT LOG
Every call, on every plan
05 · WORKS WITH

Use Mercury from any agent.

Claude
ChatGPT
Cursor
Codex
VS Code
OpenAI Agents SDK
Claude Agent SDK
Vercel AI SDK
Mastra
LangGraph
07 · FAQ

Mercury and Arc0, answered.

Q01

Can I use Mercury with Claude, ChatGPT or Cursor?

Yes. Connect Mercury to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the Mercury actions you allow.

Q02

How do users connect Mercury?

Mercury runs its own MCP server behind OAuth. Arc0 registers the client, users approve access on Arc0 Connect, and your agent reaches Mercury through the same endpoint, policies and audit log as every other app.

Q03

Which Mercury actions can my agent take?

30 in total: 30 read, 0 write and 0 destructive. Your policies decide which of them each agent may call.

Q04

Can I make my agent read-only in Mercury?

Yes. Allow read actions and deny writes in the Mercury policy. Your agent can still look things up, and any write it attempts is blocked and logged.

Q05

Can my own backend call Mercury too?

Yes. The same Mercury connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.

Get started

Plug Mercury into your agent.

Your users connect Mercury once, under your brand. Your agent gets 30 actions behind your policies, with every call on the record.

Free to build · MCP + REST · Audit log on every plan