More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Loan Approval AI Predictor

    1

    Predict loan approval outcomes using a trained Random Forest model and generate explainable predictions with SHAP feature analysis.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Loan Approval AI Predictor

    Loan Approval AI Predictor

    Example session with this skill installed

    Evaluate this loan application: Income 6000, Co-applicant 2000, Loan 150k, Term 360 months, Credit History 1. I need to know the decision and the main reasons.

    • Read your context and instructions
    • Compiled the loan approval ai

    The loan application has been processed using the Random Forest classifier.

    Loan Evaluation Result

    Based on the provided financial profile, the model predicts that the loan is Approved.

    MetricValue
    PredictionApproved
    Model Confidence92.4%
    Total Monthly Income8000.00
    Estimated Monthly EMI416.67
    Disposable Income Buffer7583.33

    Feature Attribution (SHAP)

    Positive Factors: The primary driver for approval is the satisfactory

    Credit_History (1), which matches historical compliance guidelines. The high

    Balance_Income (disposable buffer) significantly reduced the predicted risk.
    *

    Risk Factors: The requested

    LoanAmount is within acceptable limits, though it remains the largest single draw on the probability score compared to the income baseline.

    Next steps

    1. Review the generated waterfall plot in the output/ directory for a visual breakdown of feature weights.
    2. Verify the applicant's primary income documentation against the declared $6,000.00.
    3. Submit the model output to a certified human underwriter for final review.

    ⚠️

    Decision-Support Notice: This evaluation is generated by a machine learning model trained on historical data. It is a decision-support estimate and does not constitute a guaranteed loan approval, binding underwriting commitment, or professional financial advice.

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

    About this skill

    Financial decisions require transparency, not just a binary result. This skill provides a local, machine-learning-driven assessment for consumer loan applications, delivering both a prediction and a quantitative SHAP explanation. It eliminates the guesswork in credit scoring by identifying the specific financial drivers, like debt-to-income ratios or credit history compliance, that influence the model's decision.

    What it does

    • Loan Eligibility Scoring predicts approval or rejection outcomes using a 150-tree Random Forest classifier.
    • Interpretable Explanations generates SHAP feature attributions to show exactly why a loan was approved or rejected.
    • Financial Metric Engineering calculates monthly installment obligations (EMI) and disposable income buffers automatically.
    • Bias Mitigation excludes demographic fields like gender and marital status from the active classifier to focus on financial data.
    • Probability Analysis returns statistical confidence levels for every decision to help underwriters assess risk.

    How it works

    1. Input Validation accepts a JSON payload containing income, loan amount, term, and credit history.
    2. Prediction Execution runs the predict.py script to generate a prediction class and probability score.
    3. SHAP Attribution runs the explain.py script if a narrative justification of the result is required.
    4. Result Synthesis parses the JSON output into a structured response including financial metrics and risk factors.

    Frameworks & tools

    Built with Python 3.9+, Scikit-learn for the Random Forest model, SHAP for explainability, and Joblib for model artifact management. It uses Pandas for data handling and Matplotlib for generating waterfall plots of feature importance.

    Why this beats prompting it yourself

    Prompting an LLM to guess credit risk results in hallucinations and inconsistent logic based on vague training data. This skill uses a locally executed, trained model with specific feature weights, ensuring the math behind EMI and income buffers is exact every time.

    Use cases

    • Generating preliminary credit risk assessments for consumer loan applications.
    • Providing "Reason for Rejection" letters based on quantitative feature importance.
    • Batch processing structured applicant records for portfolio risk evaluation.
    • Calculating applicant disposable income and debt-service capacity.

    Known limitations

    High sensitivity to credit history, where a '0' value typically triggers a rejection regardless of income. Requires pre-trained .pkl artifacts and does not support commercial or mortgage underwriting.

    High empirical weight on credit history usually results in rejection if history is 0. Model artifacts are in pickle format. Not for legally binding underwriting or commercial loans.

    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 today

    Requires

    Python

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Listedtoday

    What's inside

    Frequently Asked Questions

    Unlike LLMs that may hallucinate logic, this skill uses a deterministic Random Forest model and SHAP values to provide mathematically grounded reasons (like debt-to-income ratios) for every decision.

    The purchase includes the core Python logic, the 150-tree Random Forest model artifacts (.pkl files), and the SHAP explainability modules for generating waterfall plots and feature attributions.

    This skill is designed to integrate with any AI agent or workflow that supports local Python execution, requiring Python 3.9+ and standard libraries like Scikit-learn and Pandas.

    Yes, the skill is specifically engineered to exclude demographic fields like gender and marital status, ensuring the model focuses strictly on financial metrics and credit history.

    The current model is optimized for consumer loans and is highly sensitive to credit history; it is not currently calibrated for commercial or complex mortgage underwriting.