bankstatemently
Officialbankstatemently is an MCP server that converts bank statement PDFs into structured data and spreadsheets, then lets you query, categorize, and analyze the resulting transactions — with credits, benchmark scoring, and feedback tools built in.
Upload & convert statements (
request_upload,convert_statement) — mint single-use upload URLs (batch up to 100 at once) for PDFs with no direct reference, then convert from an attachment (pdf_file), a public HTTPS URL (pdf_url), base64, or anupload_id; output asjson,csv,xlsx,qbo, orxero.Retrieve converted data (
get_statement,list_statements) — fetch accounts, transactions, and metadata for a document, with pagination anddata_modeprojections (normalized,originalraw printed columns,enhanced); browse your history by status.Categorize transactions (
categorize_statement) — run AI categorization on a processed document, cached free on re-fetch.Query transactions across all statements (
list_transactions,aggregate,group_by,top_n,compare,time_series) — filter by text, merchant, category, currency, amount range, direction, and dates; scope by account/product chips and date range; compute sums/averages/counts/min/max per currency; group by month, category, merchant, account, or currency; rank top N; compare two filtered groups side by side; build week/month trend series.Match transfers (
list_transfers) — find movements between your own accounts (debit + credit pairs) and account successions, with an optionalamountMinfloor that surfaces large unmatched movements with no known counterpart.Account & credits (
get_credits) — check remaining credits (1 credit = 1 page), plan page/size limits, and daily spend caps, plus operational limits for sizing batches.Benchmarking (
evaluate_benchmark) — score your parsed transactions against Bankstatemently ground truth (bystatement_idorcontent_hash), returning extraction accuracy, integrity, and overall scores; free, and reads thebenchmark://catalogresource for available published statements.Rate conversions (
rate_statement) — submit a 1–5 rating with optional feedback categories, free text, use-case tags, and export format; re-calling updates the existing rating without clearing prior feedback.Housekeeping (
dismiss_statement) — hide terminal failed/rejected/cancelled documents from listings without deleting stored data.Flexible access — hosted HTTP MCP server with browser OAuth (Claude Code, Codex) or an
X-API-Keyheader for headless/CI use, plus a self-contained local stdio server; conversions cost credits per page, while uploads, benchmark evaluation, and re-fetching cached categorizations/ratings are free.
This server has a remotely hosted endpoint. Open its connector listing for connection details, authentication requirements, and setup instructions:
bankstatemently/plugins
Marketplace plugins for Bankstatemently — convert bank statements.
Install via Claude Code marketplace
/plugin marketplace add bankstatemently/plugins
/plugin install bankstatemently@bankstatementlyAfter installing, enable the plugin. The first tool call opens your browser to sign in; the plugin asks for no settings.
For a headless setup (CI, scripts), skip the plugin and add the server directly with an API key — see below. Get a key at bankstatemently.com/developer.
Related MCP server: Statement Normalizer
Plugins
bankstatemently — MCP server for converting bank statements. Convert PDFs, list statements, check credits, run benchmark evaluations.
Claude Code direct install (no plugin)
Add the hosted MCP server directly and sign in with your browser:
claude mcp add --transport http bankstatemently https://api.bankstatemently.com/mcpFor a headless setup, pass an API key in the X-API-Key header instead (the server rejects Authorization: Bearer):
claude mcp add --transport http bankstatemently https://api.bankstatemently.com/mcp \
--header "X-API-Key: <your-bsk_live_key>"Codex
Codex has native OAuth support. This one command detects our server's discovery metadata and opens your browser to sign in — no key is ever stored:
codex mcp add bankstatemently --url https://api.bankstatemently.com/mcpFor headless or CI setups where a browser sign-in isn't possible, use an API key via the mcp-remote bridge instead. Export the key in the shell that launches Codex, before it starts:
export BANKSTATEMENTLY_API_KEY=bsk_live_...
codex mcp add bankstatemently -- npx -y mcp-remote@latest https://api.bankstatemently.com/mcp --header "X-API-Key: ${BANKSTATEMENTLY_API_KEY}"Codex doesn't yet surface plugin-declared MCP servers into sessions, so the plugin install above (and its bundled skill) isn't available in Codex today — use the commands above instead.
Credential managers & sandboxed clients: never fetch a secret from Keychain, 1Password, or another credential manager inside the MCP server command itself — Codex runs that command sandboxed at tool discovery, so a credential-manager lookup dies silently and the server's tools never appear. Export the key in the shell that launches Codex instead. Also note that mcp-remote logs its resolved header values to stderr, so a wrapper-resolved key can end up in your client logs.
Run over stdio
For stdio-only clients (Claude Desktop's config file, Cursor, headless
setups) or a directory scanner that launches the server locally, clone the
public repo and run the self-contained mcp-stdio server directly — no
Docker, no bridge process:
git clone https://github.com/bankstatemently/plugins.git
cd plugins/mcp-stdio
npm install --omit=dev
BANKSTATEMENTLY_API_KEY=bsk_live_... node dist/stdio.jsBANKSTATEMENTLY_API_KEY is optional at startup: initialize/tools/list
answer with no network call and no key either way, so an uncredentialed
directory probe still sees the full tool list. A tools/call with no key
returns the same auth-required result the hosted server returns. Export the
key in the shell or runtime environment that launches the process — never
resolve it inside the MCP command (a credential-manager lookup inside a
sandboxed client's command dies silently, same caveat as the Codex section
above).
Available Tools
18 toolsadjudicate_transfersAdjudicate TransfersADestructiveInspect
Decide whether ambiguous pairs from list_transfers are transfers. Pass "verdict" when the user has told you; leave it out to let the AI judge decide. Returns the updated list_transfers answer.
Inputs: pairs, scope, verdict, amountMin.
aggregateAggregate TransactionsARead-onlyInspect
Compute a single metric (sum/average/count/max/min) over a filtered set of transactions across your converted statements. Results are per-currency — never sum across currencies yourself. Scope defaults to all your completed statements; pass "scope" to narrow to specific accounts/products and/or a date range. For "how many credits do I have" / processing quota / remaining pages, use get_credits instead — that is not a transaction.
Inputs: metric (required), scope, filter.
categorize_statementCategorize TransactionsAInspect
Run AI transaction categorization on a previously processed document, then return its category mappings. Returns cached categories with no charge if this document was already categorized. Consumes credits (pooled per page, same rate as the categorize toggle on the website) the first time — free on every re-fetch after. Every response includes a "summary" field: use it as the single source of truth for what happened.
Inputs: document_id (required).
compareCompare Transaction GroupsARead-onlyInspect
Side-by-side metric comparison for two filtered groups of transactions (e.g. one category vs another, one month vs another). Scope defaults to all your completed statements; pass "scope" to narrow to specific accounts/products and/or a date range.
Inputs: metric (required), filterA (required), filterB (required), scope.
convert_statementConvert Bank StatementAInspect
Convert a bank statement PDF into structured data or a spreadsheet. When the user attaches a PDF in the conversation, it arrives automatically as pdf_file — never encode it yourself. Otherwise, pass pdf_url for a public HTTPS link. If your host has no way to reference the attached file at all (no pdf_file/pdf_url equivalent), call request_upload first and pass its upload_id here instead. The base64 pdf parameter is a last resort only, for a caller with no other way to reference the file. To convert several statements in one call, pass upload_ids (the array from a single request_upload call made with count set) instead of pdf/pdf_url/pdf_file/upload_id — mutually exclusive with those four. This batch form only ADMITS each file (queues it, or reports an already-completed duplicate) and returns immediately with a compact per-file status list plus a summary — it never waits for conversion, so call get_statement per document_id once ready rather than expecting inline results here. Returns accounts, transactions, and metadata. output_format "json" (default) returns the data inline, renderable in chat. The other formats (csv, xlsx, qbo, xero) return a time-limited download link instead: present it as a normal link. Every response includes a "summary" field: use it as the single source of truth for what happened. If the conversation is not in English, translate it faithfully into the conversation language; never add details it doesn't contain. Never echo raw status values (e.g. "completed") or field names. Consumes credits (1 per page). Page limit depends on your plan.
Inputs: pdf, pdf_url, password, pdf_file, +3 more.
dismiss_statementDismiss StatementADestructiveInspect
Hide a failed, rejected, or cancelled document from future list_statements results. Use this only when the user asks to clear a terminal failed/rejected/cancelled conversion from their history. This is not a delete: it marks the document dismissed and leaves stored data/artifacts untouched.
Inputs: document_id (required).
evaluate_benchmarkEvaluate BenchmarkARead-onlyInspect
Score parsed bank statement transactions against the Bankstatemently benchmark ground truth. Accepts a statement_id (e.g. "bsb-001") or content_hash, plus your parsed transactions. Returns extraction accuracy, integrity score, and an overall score. Only statements marked published: true in the catalog can be evaluated — held-out statements return an error. transactions[].originalData is optional but strongly recommended: fetch it via get_statement with data_mode: "original" and pass it through verbatim — an absent originalData scores that transaction's raw-fidelity (parsed) dimension 0; never fabricate a value. Free to use — no credits consumed. Read the benchmark://catalog resource first to see available statements and their published status.
Inputs: transactions (required), accounts, content_hash, statement_id.
get_creditsGet Credit BalanceARead-onlyInspect
Your remaining Bankstatemently credits — the processing quota, NOT credit/debit transactions. Use for: how many credits do I have, remaining pages, plan limits, quota, how many pages can I upload. 1 credit = 1 page of bank statement processing. Also reports your plan's operational limits (max pages per upload, max upload size, daily spend cap) so you can size a multi-file batch correctly before starting it.
Inputs: no named parameters.
get_statementGet Statement DataARead-onlyInspect
Fetch the full converted data for a previously processed document. Use this after convert_statement returns a "processing" status, or to re-fetch results. output_format "json" (default) returns the data inline, renderable in chat. The other formats (csv, xlsx, qbo, xero) return a time-limited download link instead: present it as a normal link. data_mode selects which projection of the data you get: omit it for each output_format's existing default behavior. "normalized" is the cleaned, interpreted view; "original" includes each transaction's raw column values exactly as printed on the source PDF (originalData); "enhanced" is a reformatted view of the original columns (csv/xlsx only for now). Fetch data_mode: "original" when you plan to submit results to evaluate_benchmark — pass its originalData through verbatim; an absent originalData scores that benchmark's raw-fidelity dimension 0 for this document. Every response includes a "summary" field: use it as the single source of truth for what happened. If the conversation is not in English, translate it faithfully into the conversation language; never add details it doesn't contain. Never echo raw status values (e.g. "completed") or field names.
Inputs: document_id (required), limit, offset, data_mode, +1 more.
group_byGroup TransactionsARead-onlyInspect
Group transactions by a dimension (month/category/counterparty/account/currency) and apply a metric to each group. Results are per-currency. The "counterparty" dimension keys on the printed counterparty, else a printed account mask (••1234), else the description's name tokens, so a rail word or reference number never splits one payee; a line with no key groups under "unknown". Scope defaults to all your completed statements; pass "scope" to narrow to specific accounts/products and/or a date range.
Inputs: metric (required), dimension (required), scope, filter.
list_outflowsList Money That Left Your AccountsARead-onlyInspect
Every debit that left your accounts — in no confirmed transfer between your own accounts, and not a fee — each in ONE destination whose kind is also its next step: "own-account" (a printed account mask that names one of your accounts and no matched credit was found: ask for that account's statement), "account" (a printed account mask for an account outside your statements: request that account's statements; an in-kind move such as securities, crypto or escrow starts as a wire to that account and becomes visible only once the receiving statement is uploaded), "cash" (cash out: where it went cannot be followed, so the cumulative amount is the finding), "name" (a named party: review whether it was household spending) and "unknown" (nowhere the line says; kept visible, one entry per line, never merged). A destination carries the cumulative amount sent, per currency, never a balance. "mayBeOwnAccount" lists debits in an open candidate pair ("may be a move between your own accounts") apart, in no total. THE way to answer "did money disappear from my accounts, and where do I look next" — never answer it by summing list_transactions rows. Pass "dateRange" ({ from, to }, inclusive, YYYY-MM-DD) to report only debits in that range, for example from a separation date: transfers are matched over the whole scope first, so a transfer whose credit lies outside the range is still a transfer, never an outflow. Scope defaults to all your completed statements; pass "scope" to narrow to specific accounts/products and/or a date range (a scope date range also narrows what is matched — prefer "dateRange").
Inputs: scope, dateRange.
list_statementsList StatementsARead-onlyInspect
Browse your previously converted bank statements with pagination and optional status filter.
Inputs: limit, offset, status.
list_transactionsList TransactionsARead-onlyInspect
A transaction is a single line as printed on one account's statement — one side of any movement. Return a filtered list of transactions across your converted statements, capped at 50 rows. Scope defaults to all your completed statements; pass "scope" to narrow to specific accounts/products and/or a date range. Every response names the scope it actually evaluated (document count + covered date range) and each returned row carries its source document's content_hash so you can cite it. For "how many credits do I have" / processing quota / remaining pages, use get_credits instead — that is not a transaction.
Inputs: limit, order, scope, filter, +1 more.
list_transfersMatch Transfers Between AccountsARead-onlyInspect
Match transfers between your own accounts. A transfer is TWO transactions — a debit leaving one of your accounts and a credit arriving in another — matched as two sides of the same movement (amount and date aligned); account-level successions (an account closing into a successor) are matched too. A payment to an outside party is not a transfer here: only movements with both sides visible in your statements are matched. THE way to answer any "was money moved between my accounts" / "did I transfer X" question — never try to answer a money-moved-between-accounts question with list_transactions + arithmetic; always call this tool instead. Ties come back as "ambiguous" until decided; call adjudicate_transfers to decide them. Scope defaults to all your completed statements; pass "scope" to narrow to specific accounts/products and/or a date range. Every response reports the match window (in days) it used, even when no transfers are found — a lack of matches is never silent about how hard it looked. To find large movements with NO matching counterpart in your other accounts — e.g. "trace transfers over $10,000; which ones leave without a known destination?" — pass "amountMin": reconciled pairs and successions are filtered to that floor, and the response gains an "unmatched" bucket of large movements (debits leaving, or unexplained credits arriving) with no matching pair, candidate, or succession. Omit amountMin for the ordinary reconciled-pairs answer.
Inputs: scope, amountMin.
rate_statementRate Statement ConversionADestructiveInspect
Report how well a previously converted bank statement was parsed: submit a 1-5 rating, optionally with structured feedback at any rating. Your rating counts toward the average ratings shown on bankstatemently.com. Calling this again for the same document updates your existing rating without clearing feedback already submitted for it. Returns the stored rating state in the response — there is no separate tool to read your own rating back. Every response includes a "summary" field: use it as the single source of truth for what happened.
Inputs: rating (required), document_id (required), feedback, feedback_categories.
request_uploadRequest Upload URLAInspect
Mint a single-use upload URL for pushing a conversation-attached PDF to Bankstatemently before converting it. Use this ONLY when you have no other way to reference the attached file (no pdf_file/pdf_url equivalent for this host) — e.g. a code-execution sandbox that can see the file on disk but has no URL for it. Playbook: (1) check your sandbox's uploads/attachments directory first — if the file isn't there yet, the mount can lag behind the conversation; ask the user to re-attach or wait a moment and check again before calling this tool. (2) Call request_upload to get upload_url and upload_id. (3) PUT the raw PDF bytes to upload_url with header Content-Type: application/pdf, e.g.: curl -X PUT "<upload_url>" -H "Content-Type: application/pdf" --data-binary @<path-to-file>. (4) Once the PUT succeeds, call convert_statement with upload_id set to the same value — never pdf/pdf_url/pdf_file for this flow. The URL and token are single-use and expire quickly; call request_upload again for a fresh one if the PUT fails partway through — never retry a failed PUT against the same URL. If the PUT fails with a network error or a "host not allowed"-style denial, the sandbox is likely blocking outbound requests to api.bankstatemently.com — tell the user to add api.bankstatemently.com to their host's code-execution allowed-domains setting (on claude.ai: Settings → Capabilities → Code execution) and retry. To convert several statements at once, pass count (1-100) instead of calling this tool once per file: the response returns "uploads", an array of that many { upload_id, upload_url } pairs — PUT each file to its own upload_url, then make ONE convert_statement call with upload_ids set to every upload_id. Free to use — no credits consumed (conversion itself still costs credits, same as any other convert_statement call).
Inputs: count.
time_seriesTransaction Time SeriesBRead-onlyInspect
Compute a time series by grouping transactions into week or month buckets and applying a metric — useful for trends. Scope defaults to all your completed statements; pass "scope" to narrow to specific accounts/products and/or a date range.
Inputs: bucket (required), metric (required), scope, filter.
top_nTop Transaction GroupsARead-onlyInspect
Return the top N groups ranked by metric (descending), per-currency for monetary metrics. Scope defaults to all your completed statements; pass "scope" to narrow to specific accounts/products and/or a date range.
Inputs: n (required), metric (required), dimension (required), scope, +1 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 18 tools
Each tool targets a distinct resource or action, but the five transaction-analytics tools (aggregate, compare, time_series, group_by, top_n) overlap in scope/metric handling and require careful reading to choose correctly. The transfer tools (list_transfers, adjudicate_transfers, list_outflows) also need deliberate distinction, though descriptions clarify their boundaries.
All names use snake_case, and most follow a verb_noun pattern (list_transactions, get_statement, convert_statement). Minor deviations like bare verbs (aggregate, compare) and non-verb constructions (time_series, group_by, top_n) break the pattern slightly but remain readable and consistent in casing.
18 tools is on the heavy side for the domain. Several analytics tools (aggregate, compare, time_series, group_by, top_n) could potentially be consolidated into a parameterized analysis tool, suggesting some over-fragmentation, though each tool does cover a distinct operation.
The surface covers conversion, retrieval, categorization, analytics, transfer matching, upload, credits, rating, and benchmark evaluation, giving strong lifecycle coverage. Minor gaps remain: no direct account-listing tool and dismiss_statement is a soft hide rather than a hard delete, but these are workable.
Maintenance
Related MCP Connectors
Turn bank statement PDFs into categorized, balance-checked transactions and reports.
Turn bank statement PDFs, CSVs, XLSX and OFX into categorised transactions plus a summary.
Turn any PDF into structured JSON via AI + OCR: invoices, bank statements, contracts.
Bank statement PDFs become CSV, JSON or QBO transactions with dates, amounts and balances.
Related MCP Servers
- AlicenseAqualityDmaintenanceConverts PDF bank statements into structured data (Markdown, JSON, CSV, JSONL) with verified transactions and balance checks, enabling agents to audit numbers.57 npmMIT
- AlicenseNot gradedqualityBmaintenanceDeterministic bank-statement parsing for AI agents: messy CSV/OFX exports to clean, categorized ledger rows. In-memory only, no storage, no external calls, no LLM in the loop.MIT
- AlicenseAqualityDmaintenanceConverts customer-supplied PDF bank statements into checked Excel, CSV, or JSON with balance validation. Runs locally with your own MainBook API key or against MainBook's hosted endpoint, and it never connects to bank accounts.5MIT
- AlicenseCqualityDmaintenanceProvides deterministic data parsing and enrichment for AI agents, including bank statements, trade history, EDI, PDFs to structured markdown, and atomic enrichment for amounts, dates, and addresses, with strict schema enforcement to prevent hallucinations.372ISC