The field of software development and blockchain technology is moving towards enhancing security and reliability. Researchers are exploring the use of Large Language Models (LLMs) to detect and repair bugs and security vulnerabilities in code, with promising results in identifying syntactic and semantic issues. However, the performance of LLMs diminishes when dealing with complex security vulnerabilities and large-scale production code. In the area of blockchain technology, smart contract security is a growing concern. New tools and techniques are being developed to analyze and identify security risks in smart contracts, including static analysis tools and heterogeneous graph mining approaches. These innovations have shown significant improvements in detection accuracy and have the potential to enhance the resilience of multi-chain ecosystems. Additionally, the security of Trusted Execution Environments (TEEs) and TEE containers is being investigated, with a focus on identifying and addressing trust boundary vulnerabilities. Noteworthy papers include: LLM-GUARD, which presents a systematic evaluation of LLMs in detecting software bugs and security vulnerabilities. MoveScanner, which introduces a static analysis tool for identifying security vulnerabilities in Move smart contracts. BridgeShield, which proposes a detection framework for securing cross-chain bridge applications via heterogeneous graph mining. Characterizing Trust Boundary Vulnerabilities in TEE Containers, which analyzes the isolation strategies employed by TEE containers and identifies critical design and implementation flaws.