Server data from the Official MCP Registry
Ask your AI about bank accounts, spending, debts, holdings, and investment activity.
Ask your AI about bank accounts, spending, debts, holdings, and investment activity.
Remote endpoints: streamable-http: https://api.openbudget.sh/mcp
Valid MCP server (1 strong, 1 medium validity signals). No known CVEs in dependencies. Imported from the Official MCP Registry.
Endpoint verified · Requires authentication · 1 issue found
Security scores are indicators to help you make informed decisions, not guarantees. Always review permissions before connecting any MCP server.
This plugin requests these system permissions. Most are normal for its category.
Remote Plugin
No local installation needed. Your AI client connects to the remote endpoint directly.
Add this to your MCP configuration to connect:
{
"mcpServers": {
"sh-openbudget-mcp": {
"url": "https://api.openbudget.sh/mcp"
}
}
}From the project's GitHub README.
Your bank accounts, spending, debts, and investments are available to your AI assistant.
OpenBudget is a hosted Model Context Protocol (MCP) server that gives Claude, ChatGPT, Cursor, Codex, and other MCP clients access to your financial data. That includes transactions, balances, net worth, credit cards, loans, mortgages, investment holdings, and investment activity synced from US institutions through Plaid. You can also add physical cash, gift-card balances, real estate, private-company interests, and other assets you track yourself.
You don't run anything. The server is hosted on OpenBudget's infrastructure. You add one URL to your client, sign in once, and start asking.
https://api.openbudget.sh/mcp
Plaid connections provide read-only financial data. OpenBudget never stores your financial-institution credentials, and your AI assistant cannot move money or place trades. Every write stays inside your OpenBudget account: categories, notes, rules, manual accounts, custom holdings, and custom activity.
Pick your client below, or see the full step-by-step guides at openbudget.sh/connect-ai.
openbudget.sh, paste https://api.openbudget.sh/mcpFull guide: openbudget.sh/connect-ai/claude
claude mcp add --transport http openbudget https://api.openbudget.sh/mcp
Add --scope user to make it available in every project. Then run /mcp inside a session, select openbudget, and choose Authenticate.
Full guide: openbudget.sh/connect-ai/claude-code
OpenBudget, paste the URL, select OAuthFull guide: openbudget.sh/connect-ai/chatgpt
Add to ~/.cursor/mcp.json (global) or .cursor/mcp.json (project):
{
"mcpServers": {
"openbudget": {
"url": "https://api.openbudget.sh/mcp"
}
}
}
Then enable the server under Cursor Settings → Tools & MCP → Connect.
Full guide: openbudget.sh/connect-ai/cursor
codex mcp add openbudget --url https://api.openbudget.sh/mcp
codex mcp login openbudget
Or add to ~/.codex/config.toml:
[mcp_servers.openbudget]
url = "https://api.openbudget.sh/mcp"
Full guides: codex · codex-cli
Add a remote / custom MCP server over the Streamable HTTP transport, pointing at https://api.openbudget.sh/mcp. Most clients also accept the mcpServers JSON shape shown in the Cursor section above. The first call opens your browser to sign in and approve access.
Full guide: openbudget.sh/connect-ai/other
The server ships with 22 tools, each shown with a live demo at openbudget.sh/tools. You rarely call them by name. Ask in natural language and your client picks the right one.
| Tool | What it does |
|---|---|
list_accounts | Every linked account with balances, institution, and a normalized net-worth view. |
search_transactions | Search and filter transactions by text, category, account, date range, or amount, with settlement links and merchant location details when Plaid supplies them. |
get_liabilities | Credit cards, student loans, and mortgages: APRs, due dates, minimum payments, payoff data. |
get_liability_history | How those liabilities changed over time, including APR moves, balance trends, and YTD interest. |
list_available_categories | The full transaction category taxonomy. |
list_categorization_rules | Your auto-categorization rules, in evaluation order. |
update_transaction | (write) Recategorize a transaction, add a note, split it, or hide it from searches. |
create_transaction | (write) Log an off-bank purchase made with physical cash, a gift card, or store credit. |
delete_transaction | (write) Remove a manually-created transaction (bank transactions are hidden with update_transaction instead). |
upsert_manual_account | (write) Create or update a manual off-bank account such as a cash wallet or gift-card balance. |
save_categorization_rule | (write) Create a rule once, such as "payments to our nanny are childcare," and new transactions follow it. |
delete_categorization_rule | (write) Remove a rule. |
update_liability | (write) Note or manual balance override on a credit card or loan. |
Investor includes every Base tool and adds the tools below.
| Tool | What it does |
|---|---|
search_investment_holdings | Filter connected and custom positions by account, ticker, name, category, source, or closed status, then sort by value, quantity, name, or ticker. Returns the position data your AI needs to calculate allocation, gains, and concentration. |
search_investment_transactions | Search buys, sells, dividends, interest, contributions, withdrawals, fees, taxes, property activity, and adjustments. |
list_available_investment_categories | Return the separate category taxonomies for holdings and investment activity. |
update_investment_holding | (write) Categorize a holding, change its display fields, or manage its notes. Custom holdings also support value, quantity, price, cost basis, ticker, currency, and status changes. |
update_investment_transaction | (write) Categorize investment activity, add a note, or ignore it. Custom activity also supports amount, date, quantity, price, fees, ticker, and currency changes. |
create_investment_holding | (write) Add real estate, private-company interests, collectibles, offline positions, or other custom assets. |
delete_investment_holding | (write) Delete a custom holding. Connected holdings remain synced from the institution. |
create_investment_transaction | (write) Add property purchases, distributions, private contributions, fees, valuation adjustments, and other custom investment activity. |
delete_investment_transaction | (write) Delete custom investment activity. Connected activity can be ignored instead. |
Transaction results can include pendingTransactionId, which links a posted transaction to its earlier pending record, and a structured location with address, city, region, postal code, store number, latitude, and longitude. These fields make receipt reconciliation reliable when a tip or pre-authorization changes the settled amount.
More in examples/prompts.md.
Bank, brokerage, and retirement connections go through Plaid with read-only data access. OpenBudget never sees or stores your financial-institution credentials. Your AI assistant cannot move money, initiate transfers, place trades, or modify anything at your institution. Every write stays inside your OpenBudget account: transaction and investment categories, notes, ignored flags, splits, rules, liability overrides, manual accounts, custom holdings, and custom activity. Every tool declares MCP annotations (readOnlyHint, destructiveHint) so clients can show what each call can do.
OAuth grants your AI client access to your own financial data in OpenBudget. See the Terms of Service, Privacy Policy, and Security pages.
The configuration snippets and documentation in this repository are MIT-licensed.
The OpenBudget MCP service is hosted and governed by its Terms of Service. The MIT license covers the configuration snippets and documentation in this repository. This repository does not contain the server's source code.
Be the first to review this server!
by Toleno · Developer Tools
Toleno Network MCP Server — Manage your Toleno mining account with Claude AI using natural language.
by mcp-marketplace · Developer Tools
Create, build, and publish Python MCP servers to PyPI — conversationally.
by Microsoft · Content & Media
Convert files (PDF, Word, Excel, images, audio) to Markdown for LLM consumption