Recent developments in financial systems and the experience of banking crises have shown that credit risk is not merely the result of individual borrower defaults; through contagion mechanisms and interdependencies within credit portfolios, it can evolve into a source of systemic risk. In Islamic banking, the diversity of financing contracts and their heterogeneous risk characteristics further complicate credit risk management, highlighting the need for a network-based and systemic approach. Accordingly, this study aims to design a credit risk management policy for bank customers by identifying and analyzing risk contagion mechanisms among different financing contracts in the credit portfolio of Bank Saderat Iran. To achieve this objective, daily data on the ratio of non-performing loans to total facilities for four major Islamic financing contracts—Murabaha, Joalah, Mudarabah, and Civil Partnership—during 2011–2021 were analyzed using Conditional Value-at-Risk (CoVaR), Vector Autoregression (VAR), Dynamic Conditional Correlation Multivariate GARCH (DCC-MGARCH), and BEKK models. The results indicate that credit risk within the bank’s portfolio exhibits a dynamic, networked, and asymmetric structure, with credit shocks transmitted through distinct contagion channels across contracts. The DCC-MGARCH estimates confirm significant time-varying conditional correlations, as the shock-response parameter (0.81) substantially exceeds the persistence parameter (0.188), indicating the dominant role of newly emerging shocks in shaping systemic risk. The BEKK results further reveal that Mudarabah acts as the primary transmitter of risk within the portfolio, generating the strongest spillovers to Civil Partnership (0.767), Murabaha (0.605), and Joalah (0.328). Conversely, Civil Partnership displays the highest credit risk volatility and is the most vulnerable contract to incoming shocks. Based on these findings, the study develops a policy framework consisting of four complementary instruments: concentration limits on contracts with systemic risk transmission capacity, enhanced collateral requirements for customers simultaneously engaged in Mudarabah and Civil Partnership contracts, risk-sensitive allocation of economic capital and loan-loss reserves according to each contract’s position in the contagion network, and a CoVaR-based early warning mechanism for the timely identification and containment of emerging credit shocks. The main contribution of this study is moving beyond conventional risk measurement toward the design of quantitative and operational policy tools for credit risk management from a contagion-network perspective in Islamic banking.