Overview

US bank data from the FDIC BankFind Suite API, which is public and needs no API key. Find any FDIC-insured bank or savings institution, including closed, merged, and failed ones, then read its quarterly Call Report financials back to 1984, rank it against peers, screen many banks at once, search failures since 1934, and map branch deposits and market share. Runs as a stdio process or a local Streamable HTTP server.

Institutions are keyed by their FDIC certificate number (CERT), which survives renames and charter conversions. Credit unions are insured by the NCUA and are not in this data.

Tools

Dataframe workflow

1. fdic_query_financials and fdic_get_deposits return an inline preview. When the full result is larger than limit, they also stage it as a df_XXXXX_XXXXX table and name it in the response's dataset field.
2. Pass that name to fdic_dataframe_describe for the table's columns and the unit and basis of each numeric column.
3. Query it with fdic_dataframe_query: joins, aggregates, window functions, and CTEs across every staged table. register_as saves a result as a new table.
4. Tables expire after FDIC_DATASET_TTL_SECONDS (default 24 hours), and the live tables hold 1,000,000 rows together: past that, the oldest are dropped to make room. With FDIC_DATAFRAME_DROP_ENABLED=true, fdic_dataframe_drop removes one sooner.

Staging runs on DuckDB and is on by default; CANVAS_PROVIDER_TYPE=none turns it off. The Claude Desktop .mcpb bundle ships without DuckDB's native binding, so there the dataframe tools report an install hint and every other tool works normally.

Capability reference

fdic_search_institutions _tool

• Filters: name (up to 100 characters; every word must match a current, former, or trade name), up to 50 certs, state, city, bank_classes, min_assets / max_assets (thousands of US dollars), holding_company_rssd
• status is active, inactive, or any; it defaults to any when name or certs is given and active otherwise, and the applied value comes back as status_filter
• Up to 100 per page (default 20), paged with offset / next_offset; records carry active, successor_cert for merged or failed banks, holding_company.rssd, and matched_on when a former or trade name matched; requested CERTs with no record land in missing_certs

fdic_get_institution_financials _tool

• cert required; up to 25 metrics (default: a 15-metric health set); quarters 1–200 (default 8), optionally bounded by from_date / to_date as 2025-03-31, 20250331, or 2025Q1
• Rows run most recent first, each a values map of metric → number, null when not reported; metric_definitions gives each metric's FDIC field, unit, and basis
• quarters_available counts the quarters in the window before the cap; an unknown CERT fails as cert_not_found

fdic_compare_peers _tool

• cert required; report_date defaults to the latest published quarter (report_date_defaulted); up to 20 metrics
• Peer group: peer_asset_band (same by default, any, under_100m, 100m_1b, 1b_10b, 10b_250b, over_250b) plus an optional peer_state, or an explicit peer_certs list of up to 200, which can't be combined with the band or state (conflicting_peer_filters)
• Each comparison returns the institution's value, peer_median, peer_p25 / peer_p75, peer_min / peer_max, percentile, and rank of rank_of, over the peer_count_with_value peers that reported the metric

fdic_query_financials _tool

• Filters: up to 100 certs, state, min_assets / max_assets, and up to 5 metric_filters (metric with min and/or max in the metric's unit); up to 30 metrics; from_date / to_date, defaulting to the latest published quarter only
• Inline preview of up to 500 rows (default 50), ordered by sort_by / sort_order; a panel larger than the preview is staged whole as a dataframe
• The panel is capped at FDIC_PANEL_MAX_ROWS (default 50,000) and keeps the newest quarters when it binds; total_matching, rows_fetched, and panel_truncated report what was left out

fdic_search_failures _tool

• Filters: name, up to 50 certs (events before 1977 carry none), state, from_date / to_date (YYYY-MM-DD), methods, min_assets; resolution is failure by default, or assistance or all
• summary totals every matching event, not just the page, with a by_method breakdown; group_by (year, state, method, insurance_fund) adds groups
• estimated_loss_missing_count counts events without an FDIC loss estimate, and a loss total is null when none has one; sort by date, loss, or assets, up to 200 per page (default 25)

fdic_get_deposits _tool

• cert alone returns the institution's branches and its deposit share in each state (mode: "institution"); a geography (state, county, city, zip, msa_code) ranks every institution in that market with hhi (market); both give the institution's position in the market (institution_in_market). county and city require state
• year from 1994 on, defaulting to the latest survey (year_defaulted); deposits are domestic branch deposits as of June 30, in thousands of US dollars
• Up to 200 branches or ranked institutions inline (default 25); a larger set is staged as a dataframe, and total_rows gives the full count

fdic_list_reference _tool

• One topic per call: metrics, bank_classes, failure_methods, insurance_funds, peer_asset_bands, or coverage
• metrics lists all 49 metric names with FDIC field, unit, basis, caveat, and in_default_set; served from built-in tables with no request to FDIC

fdic_dataframe_describe _tool

• name (a df_XXXXX_XXXXX from a dataset field) returns that table in full: source tool, the parameters it was called with, row count, expiry, column_schema, and column_units
• Without name, lists the live tables newest first, 50 per page, paged with offset / next_offset. Over HTTP with MCP_AUTH_MODE=none, where every caller shares one canvas, the listing is off and fails as listing_unavailable; a table is reached by its name
• Fails as canvas_unavailable when dataframe staging is off