August 2025
Rsdoctor 1.2 has been released! 🎉
During the Rspack build process, certain modules are hoisted or aggregated into a single closure, forming a concatenated module to improve browser execution efficiency and reduce artifact size. Previously, Rsdoctor could not further decompose and analyze the internal structure of these concatenated modules, as they cannot be further split through AST parsing.
Rsdoctor 1.2 supports the ability to analyze concatenated module sizes, helping developers accurately understand the actual build size of sub-modules (aggregated modules) after tree shaking and compression, facilitating analysis of how concatenated modules affect final bundle size and optimization of code splitting strategies.
Additionally, the Rsdoctor plugin built into Rspack (>=1.4.11) has enhanced source map capabilities, allowing seamless analysis of concatenated modules without enabling source maps. However, Webpack projects still require source maps to be enabled.
Previously, Rsdoctor's Treemap view was implemented based on webpack-bundle-analyzer, which required Rsdoctor to go through webpack-bundle-analyzer's processing pipeline again after completing its analysis, reducing overall analysis efficiency. Additionally, Treemap page loading was slow, while Treemap is precisely the most commonly used visualization view for developers when analyzing bundles.
Rsdoctor 1.2 introduces a new classic Treemap artifact analysis view, helping developers more intuitively visualize and analyze bundle composition, Assets, and Modules proportions. You can also search for module resources, click on the module resource, and zoom in to the module area.
To more accurately reflect production environment size performance, Rsdoctor has added support for analyzing gzip compressed sizes, which can be viewed on the Bundle Size page and TreeMap page, as shown below:
This shows a comparison between original size and gzip compressed size, providing more accurate reference data for production environment optimization.
Rsdoctor provides rich build analysis data, but developers need to spend time on page interactions and learning costs to perform build analysis and optimization. Therefore, we hope to leverage LLM for more intelligent build analysis to help users obtain analysis results more quickly.
Rsdoctor v1.1 introduced MCP support, which is based on Model Context Protocol (MCP) protocol, combining Rsdoctor's analysis capabilities with LLM's intelligent understanding abilities. Through natural language Q&A interactions, developers can quickly obtain build analysis results without needing to deeply understand complex analysis interfaces and data structures. Its main features include obtaining artifact information, dependency analysis, optimization suggestions, compilation performance, and tree shaking analysis among other core analytical capabilities.
Rsdoctor MCP supports natural language Q&A - you can directly ask questions like "Which packages have the largest volume?" or "Why wasn't this module tree-shaken?" The system will intelligently analyze and provide optimization suggestions, helping you quickly identify and resolve build issues.
In addition, the 1.1-1.2 versions also included other capability changes. For complete update details, please refer to: Release page