June 2026 highlights for opensumi/core focused on empowering users with configurable search behavior and stabilizing the ACP-based workflow, delivering business value through improved control, reliability, and maintainability. Key feature delivered: - Added search.followSymlinks: Introduced a new user preference to control whether symbolic links are followed during search operations. Default is enabled. The feature is wired through search services and integrated into the UI, with UI config support and i18n translations. The implementation ensures the setting is respected across content search, file search, and the top-level search pipelines, including CLI/extension paths that leverage ripgrep. Major bugs fixed: - Resolved critical compile-time errors and dependency injection issues uncovered during ACP component refactor, improving build stability. - Fixed CI/config problems that could cause flaky pipelines and delayed releases. - Hardened session handling and UI safeguards (e.g., input disabling during session loading, automatic session creation on load failures) to improve runtime stability for ACP workflows. Overall impact and accomplishments: - Business value: more precise search results and safer search behavior with user-configurable symlink handling. This reduces user confusion and increases productivity when dealing with large codebases. - Technical stability: stabilized ACP integration, improved modularization, and established groundwork for multi-workspace support and dynamic component resolution. - Developer experience: added comprehensive tests, documentation for ACP components, and alignment with contribution points to simplify future ACP enhancements. Technologies/skills demonstrated: - TypeScript, React, and DI patterns in a large modular codebase. - Cross-service integration: content-search, file-search, and core search logic with a common preference flag. - ACP architecture and multi-workspace support: component refactors, contribution points, i18n, and robust session handling. - Testing, CI, and localization practices (mocking, i18n translations) to improve quality and deliverability.