Low-Rank Adaptation (LoRA) is now the dominant parameter-efficient fine-tuning method for large language models, but the resulting open marketplace of adapters has become a supply-chain attack surface: backdoored adapters merge silently with arbitrary task adapters while preserving benign utility, and existing centralized integrity tools require marketplaces to share raw weights, which is incompatible with proprietary fine-tuning. We present FedLoRAGuard, a federated dynamic-graph framework that verifies adapter integrity without weight sharing. The LoRA ecosystem is modeled as a heterogeneous continuous-time dynamic graph with four node types and six edge types; a federated dynamic GNN is trained per-client with client-update-level (ε, δ)-DP Gaussian noise, secure aggregation, and FLTrust scoring. We prove a sensitivity bound for federated heterogeneous-GNN client updates and a certified poisoning-radius theorem that converts the per-round inference-time smoothing budget ε_r into an analytical bound on the size of a bounded-update colluding coalition k*. On 13,500 benign and backdoored adapters across four base-model families and ten attack types, FedLoRAGuard obtains 96.4% macro-F1 and a certified coalition bound k* = 25 from N = 50 marketplaces at the inference-time smoothing budget ε_r = 0.5, and empirically tolerates label-flipping Byzantine coalitions up to ρ ≤ 0.33 without certificate degradation. The DP-SGD training-time mechanism, calibrated at σ = 1.1 and Poisson subsampling q = 0.2 over T = 100 rounds, achieves a cumulative privacy budget ε_T ≈ 13.6 at δ = 10⁻⁵ under the Opacus PRV accountant, with sub-second per-adapter verification latency on A100-class hardware. The full DP + secure-aggregation stack adds a measured median per-round wall-clock overhead of ≈ 4% (pooled median across three runs of 20–60 measurement rounds on identical CPU hardware) over non-private FedAvg; the released runtime ships the instrumentation so operators can re-measure in place on their target hardware.