<![CDATA[Before the emergence of modern currency systems, human economic transactions were conducted through barter, the direct exchange of goods or services. However, contemporary barter faces persistent operational friction, including the difficulty of establishing a double coincidence of wants, the absence of standardized exchange values, and inefficient manual matching. To address these challenges, this study develops a responsive, token-free web barter platform prototype engineered using the Waterfall development model. The system utilizes Node.js and the Fastify backend framework for high-throughput, non-blocking asynchronous request processing, paired with an EJS and Webix presentation layer and a PostgreSQL relational database. To streamline partner discovery, the platform couples PostgreSQL full-text search and trigram similarity matching (pg_trgm) for typo tolerance with an automated recommendation pipeline driven by recent search history. To mitigate cold-start friction for new users, the recommendation mechanism incorporates an adaptive fallback strategy that prioritizes listings by visit frequency. User data security is enforced through HTTP-only JSON Web Token (JWT) session handling, bcrypt cryptographic password hashing, and strict AJV schema validation. Functional verification via Black-Box testing achieved a 100% success rate across 16 transactional scenarios, while an OWASP ZAP active scan detected zero high-risk vulnerabilities. Performance benchmarking with Grafana k6 confirmed robust backend efficiency, recording a 95th percentile response time of 205.58 ms under sustained multi-user load and maintaining operational resilience up to 200 concurrent virtual users. This throughput stability confirms the efficacy of persistent database connection pooling and asynchronous thread execution under concurrent traffic. Finally, usability evaluation using the System Usability Scale (SUS) with 17 respondents yielded a composite score of 71.91, classifying the platform as Good with an Acceptable level of user reception.]]>
