Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    commercial loan underwriting assistant

    1

    Commercial credit analysis fails when manual math errors, inconsistent source referencing, or overlooked policy exceptions enter the credit memo.

    $10

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    commercial loan underwriting assistant

    commercial loan underwriting assistant

    Example session with this skill installed

    Analyze this $2M loan request at 6.5% for 10 years. Borrower EBITDA is $450k with $50k capex. Collateral is a building appraised at $2.8M with a 20% haircut. E1 is the 2023 Tax Return.

    • Read your context and instructions
    • Compiled the commercial loan underwriting

    I have processed the loan request and historical data. The analysis identifies sufficient repayment capacity but highlights a specific collateral coverage gap after haircuts.

    Credit Underwriting Analysis — Borrower Corp, 2024-05-20

    1. Request, sources and assumptions

    • Loan Request: $2,000,000 at 6.50% interest, 10-year term, amortizing.
    • Sources: E1 (2023 Tax Return).

    Assumptions: Interest rate and term are user-supplied.

    Analytical estimate — validation required for default maintenance capex if not verified by E1.

    2. Application completeness and information gaps

    • Application Status: completed_with_warnings

    Information gap — organizational validation required: Detailed breakdown of 2023 cash taxes was not provided; CFADS derived from EBITDA minus Capex.

    3. Repayment capacity

    PeriodKindCFADSPro forma DSCRSurplus
    2023Historical$400,0001.48x$129,240
    • Debt / EBITDA: 4.44x
    • Loan-to-Value: 71.4%

    4. Collateral: appraised versus realizable value

    ItemStatusPledgedAppraisedRealizableNet recoverable
    Real EstatePledged$2,800,000$2,800,000$2,240,000$2,240,000
    • Shortfall vs Loan: None ($240,000 surplus over loan amount).

    5. Stress scenarios

    ScenarioStressed DSCRLiquidity headroomStressed LTVStressed coveragePolicy breaches
    Rate +2%1.34x$92,00071.4%112%None

    6. Alternative structures for analyst review

    • Interest Only: 12-month IO period would increase DSCR to 3.08x for year one.

    7. Exceptions and conflicting figures

    • Policy Check: DSCR 1.48x exceeds policy minimum of 1.25x.
    • Conflicts: None detected between E1 and request.

    8. Human approval queue and underwriting narrative

    • Marker: Repayment capacity is supported by E1 tax returns.
    • Marker: Collateral realizable value provides 1.12x coverage.
    • Management decision required: Final confirmation of maintenance capex figures to validate CFADS.

    This analysis supports, and does not replace, professional judgment and the credit authority's decision.

    Next steps

    • Verify the maintenance capex figures to finalize the CFADS calculation.
    • Review the 20% recovery haircut against current local market conditions for commercial real estate.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Commercial credit analysis fails when manual math errors, inconsistent source referencing, or overlooked policy exceptions enter the credit memo. This skill automates the rigorous preparation of commercial loan underwriting packages, ensuring every figure is auditable and every gap is flagged.

    What it does

    • Financial modeling calculates CFADS, DSCR, and leverage using a deterministic engine for exact decimal precision.
    • Collateral valuation compares pledged, appraised, and realizable values including recovery haircuts and prior liens.
    • Stress testing runs liquidity scenarios and pro forma debt service coverage under varied interest rate or cash flow shocks.
    • Audit trail construction maps every data point to E-numbered sources for transparent, evidence-based underwriting.
    • Gap detection identifies missing appraisals, document omissions, and conflicting figures across borrower statements.

    How it works

    1. Material review parses borrower statements, loan requests, and appraisals to build a structured data map with source citations.
    2. Data validation constructs a request.json file, verifying required fields like loan amount, interest rates, and historical cash flows.
    3. Engine execution runs a dedicated Python script to compute metrics without human rounding errors or manual adjustments.
    4. Report generation produces a structured credit analysis across eight specific sections, including an approval queue for the human analyst.

    Frameworks & tools

    The skill utilizes a self-contained Python calculation engine for deterministic math and follows a strict schema-driven approach for data ingestion and output.

    Why this beats prompting it yourself

    Generic prompts often hallucinate financial metrics or fail to distinguish between appraised and realizable collateral value. This skill enforces a hard separation between calculation and explanation, using a script to handle logic while maintaining a strict evidence-based audit trail that generic LLM outputs lack.

    Use cases

    • Drafting the repayment and collateral sections of a formal commercial credit memo.
    • Identifying policy exceptions and information gaps in new loan applications.
    • Running interest rate sensitivity analysis on complex commercial debt structures.
    • Reconciling conflicting figures between borrower tax returns and internal spreadsheets.

    Known limitations

    Does not perform consumer credit scoring or regulatory capital (Basel) calculations. Cannot value collateral or provide market haircuts without user-supplied inputs.

    Requires Python 3 for calculation script. Does not provide credit decisions or pricing. Missing appraisals remain unknown. One currency per request. Confirms data sharing rights before processing.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 5 days ago

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Environment Variables
    Listed5 days ago

    What's inside

    Frequently Asked Questions

    Unlike a standard LLM that might 'guess' math, this skill uses a deterministic Python engine to calculate DSCR and CFADS. This ensures decimal-perfect precision and removes the risk of financial hallucinations in your credit memos.

    The purchase includes the structured data schema for material review, the deterministic calculation scripts, and the multi-section report generation prompts. You receive the full workflow logic needed to automate the transition from raw borrower data to a structured underwriting draft.

    This skill is designed for any AI agent or platform that supports Python code execution (such as ChatGPT with Advanced Data Analysis or custom AutoGPT agents). It requires a system capable of handling structured JSON data for the input and output validation.

    The assistant creates a 'Source-to-Statement' map, linking every financial figure to an E-numbered source (e.g., E1: Tax Return). This makes it easy for senior underwriters or auditors to hover over a value and verify the exact document it originated from.

    While it automates the math and drafting, the skill includes a specific 'Approval Queue' in its output. This highlights missing documents, policy exceptions, and conflicting data points that a human analyst must address before final submission.

    The skill focuses strictly on commercial underwriting (CFADS, LTV, Stress Testing); it does not handle consumer credit scoring or regulatory capital (Basel) calculations. It also requires user-provided inputs for collateral values and market haircuts to function.