This page is a reference for the MCP tools provided by this server. All tools work in any MCP client — Claude Desktop, ChatGPT, Gemini CLI, GitHub Copilot CLI, and others.

For guided multi-tool workflows available in Claude Code, see Skills. For help choosing between tools and skills, see Choose a Workflow.

Search and Lookup Tools

Tool Use It When Notes
search ChatGPT needs citation-friendly retrieval results across institutions, failures, branches, or schema docs Standard MCP retrieval shape for ChatGPT company knowledge and deep research compatibility
fetch ChatGPT needs full citation text for an item returned by search Accepts ids such as institution:<CERT>, failure:<CERT>, branch:<id>, and schema:<endpoint>
fdic_search_institutions You need institution search results by name, state, status, asset size, or other details Good default entry point for bank discovery
fdic_get_institution You already know the FDIC CERT and need one institution record Best for direct bank lookup
fdic_search_failures You need failed-bank records, failure dates, costs, or resolution details Use when the question is specifically about bank failures
fdic_get_institution_failure You already know the failed institution CERT and want a direct lookup Best for one failed-bank record
fdic_search_locations You need branch or office locations for one bank or a geography Useful for office-level footprint questions
fdic_search_history You need structural-change records such as mergers or name changes Use when the question is about institution history

Financial and Deposit Datasets

Tool Use It When Data Cadence
fdic_search_financials You need quarterly Call Report data Quarterly — identified by REPDTE (YYYYMMDD). Quarter-end dates: 0331, 0630, 0930, 1231.
fdic_search_summary You need annual summary data Annual — identified by YEAR. Different from branch-level SOD data.
fdic_search_sod You need Summary of Deposits branch-level deposit data Annual — as of June 30 each year. Branch-level, not institution-level.
fdic_search_demographics You need quarterly demographics or market-structure fields, including office counts Quarterly — useful for office-count comparisons and geography context.

All financial amounts are in thousands of dollars unless otherwise noted.

Analysis Tools

Tool Use It When Notes
fdic_compare_bank_snapshots You want to compare multiple banks across two quarterly report dates or over a quarterly time series Best for growth, profitability, branch-count, and trend analysis
fdic_peer_group_analysis You want to benchmark one institution against a peer group at a single report date Best for ranking a bank against comparable institutions. Auto-derives peer criteria from the subject’s asset size and charter class.
fdic_qbp_lite_data You want chart-ready data for a concise QBP Lite report Aggregates public quarterly BankFind financials into executive snapshot metrics, trend series, portfolio credit-quality data, capital ratios, community-bank comparisons, and explicit exclusions for non-reproducible QBP items.

Health and Risk Analysis Tools

Tool Use It When Notes
fdic_analyze_bank_health You want a CAMELS-style health assessment for a single institution public_camels_proxy_v1 assessment with overall band, PCA capital categorization, management overlay, and trend analysis
fdic_show_bank_deep_dive You want a ChatGPT app dashboard for a single institution Returns dashboard-ready structured content and links to the embedded Bank Deep Dive widget
fdic_compare_peer_health You want to rank a group of institutions by health scores Ranks peers by proxy scores with percentiles and outlier flags for the subject institution
fdic_detect_risk_signals You want to scan institutions for early warning indicators Flags critical and warning-level issues with standardized signal codes

These tools produce analytical assessments based on public financial data. They are not official regulatory CAMELS ratings or confidential supervisory conclusions.

Credit, Funding, and Securities Detail Tools

Tool Use It When Notes
fdic_analyze_credit_concentration You want loan portfolio composition and CRE concentration risk Flags CRE and construction concentrations per interagency guidance thresholds (300% / 100% of capital)
fdic_analyze_funding_profile You want deposit composition, wholesale funding reliance, and liquidity metrics Flags brokered deposit reliance, low core deposits, and FHLB dependence
fdic_analyze_securities_portfolio You want securities holdings breakdown and concentration risk Flags high securities-to-assets, securities-to-capital, and MBS concentration

UBPR-Equivalent Analysis

Tool Use It When Notes
fdic_ubpr_analysis You want a comprehensive ratio report with summary, loan mix, capital, liquidity, and growth rates Computes UBPR-equivalent ratios from Call Report data. Not official FFIEC UBPR output.

Market Share and Franchise Tools

Tool Use It When Notes
fdic_market_share_analysis You want deposit market share and HHI concentration for an MSA or city market Requires a numeric MSABR code or city + state. Uses annual SOD data. Use fdic_search_sod to look up MSABR codes.
fdic_franchise_footprint You want one institution’s branch network across all its markets Shows deposit totals and branch counts grouped by MSA from SOD data

Holding Company Tools

Tool Use It When Notes
fdic_holding_company_profile You want all subsidiaries under a holding company with aggregated financials Look up by HC name or any subsidiary CERT

Regional Economic Context

Tool Use It When Notes
fdic_regional_context You want macro/regional economic context for a bank’s operating environment Uses FRED data for unemployment and rate trends. Gracefully degrades if FRED is unavailable.

Data Basis Reference

Dataset Cadence Date Field Dollar Units
Call Reports (financials, demographics) Quarterly REPDTE (YYYYMMDD) Thousands
Summary data Annual YEAR Thousands
Summary of Deposits (SOD) Annual (June 30) YEAR Thousands
Institution profiles Current Thousands
Failure records Event-based FAILDATE Thousands

Publication lag: FDIC data is typically available approximately 90 days after the reporting period.

Do not mix quarterly and annual data in a single prompt without acknowledging the different date bases. For more on date handling, see the Prompting Guide.