[{"data":1,"prerenderedAt":4},["ShallowReactive",2],{"readme:serena":3},"\u003Cp align=\"center\">\n  \u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Foraios\u002Fserena\u002FHEAD\u002Fresources\u002Fserena-logo.svg#gh-light-mode-only\" \u002F>\n  \u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Foraios\u002Fserena\u002FHEAD\u002Fresources\u002Fserena-logo-dark-mode.svg#gh-dark-mode-only\" \u002F>\n\u003C\u002Fp>\u003Ch3>\n    The IDE for Your Coding Agent\n\u003C\u002Fh3>\u003Cdiv align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fdiscord.com\u002Finvite\u002FcVUNQmnV4r\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fdiscord-join-5865F2?style=flat-square&labelColor=0a0e14&logo=discord&logoColor=5865F2\" alt=\"discord\" \u002F>\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Foraios\u002Fserena\u002Fmain\u002FLICENSE\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Flicense-MIT-b0e8ff?style=flat-square&labelColor=0a0e14\" alt=\"license\" \u002F>\u003C\u002Fa>\n\u003C\u002Fdiv>\n\u003Cbr \u002F>\u003Cul>\n\u003Cli>Serena provides essential \u003Cstrong>semantic code retrieval, editing, refactoring and debugging tools\u003C\u002Fstrong> that are akin to an IDE's capabilities,\noperating at the symbol level and exploiting relational structure.\u003C\u002Fli>\n\u003Cli>It integrates with any client\u002FLLM via the model context protocol (\u003Cstrong>MCP\u003C\u002Fstrong>).\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>Serena's \u003Cstrong>agent-first tool design\u003C\u002Fstrong> involves robust high-level abstractions, distinguishing it from\napproaches that rely on low-level concepts like line numbers or primitive search patterns.\u003C\u002Fp>\n\u003Cp>Practically, this means that your agent operates \u003Cstrong>faster, more efficiently and more reliably\u003C\u002Fstrong>, especially in larger and\nmore complex codebases.\u003C\u002Fp>\n\u003Cblockquote>\n\u003Cp>[!IMPORTANT]\nDo not install Serena via an MCP or plugin marketplace! They contain outdated and suboptimal installation commands. \nInstead, follow our \u003Ca href=\"#quick-start\" rel=\"nofollow ugc noopener\">Quick Start\u003C\u002Fa> instructions.\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Ch2>Quick Demo\u003C\u002Fh2>\n\u003Cp>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F8d11646e-b80e-4723-b9d7-32d6101b5f58\" rel=\"nofollow ugc noopener\">https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F8d11646e-b80e-4723-b9d7-32d6101b5f58\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>:tv: Longer video: \u003Ca href=\"https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=5QN7gN1KYLA\" rel=\"nofollow ugc noopener\">Introduction to Serena in 5 Minutes (YouTube)\u003C\u002Fa>\u003C\u002Fp>\n\u003Ch2>What Our \"End Users\" Say\u003C\u002Fh2>\n\u003Cp>While it is humans who download and set up Serena, our end users are essentially AI agents.\nAs the ones actually applying Serena's tools, they are in the best position to evaluate Serena.\u003C\u002Fp>\n\u003Cp>We crafted an unbiased evaluation prompt that leads the agent to perform ~20 routine coding tasks, \nrepresentative of everyday development work, \nin order to estimate the value added by Serena's tools when used alongside its own built-ins. \u003C\u002Fp>\n\u003Cp>Here's a one-sentence summary of what the agents had to say:\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Opus 4.6 (high) in Claude Code on a large Python codebase:\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cblockquote>\n\u003Cp>\"Serena's IDE-backed semantic tools are the single most impactful addition to my toolkit – cross-file renames, moves, and reference lookups that\nwould cost me 8–12 careful, error-prone steps collapse into one atomic call, and I would absolutely ask any developer I work with to set them up.\"\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Cp>\u003Cstrong>GPT 5.4 (high) in Codex CLI on a Java codebase:\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cblockquote>\n\u003Cp>\"As a coding AI agent, I would ask my owner to add Serena because it gives me the missing IDE-level understanding of symbols, references, and\nrefactorings, turning fragile text surgery into calmer, faster, more confident code changes where semantics matter.\"\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Cp>\u003Cstrong>GPT 5.4 (medium) in Copilot CLI on a large, multi-language monorepo:\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cblockquote>\n\u003Cp>\"As a coding agent, I’d absolutely ask my owner to add Serena because it makes me noticeably sharper and calmer on\nreal code – especially symbol-aware navigation, cross-file refactors, and monorepo dependency jumps – while I still lean\non built-ins for tiny text edits and non-code work.\"\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Cp>Different agents in different settings independently converge on the same verdict.\u003C\u002Fp>\n\u003Cp>\u003Cem>Give your agent the tools it has been asking for and add Serena MCP to your client!\u003C\u002Fem>\u003C\u002Fp>\n\u003Cp>See our \u003Ca href=\"https:\u002F\u002Foraios.github.io\u002Fserena\u002F04-evaluation\u002F000_evaluation-intro.html\" rel=\"nofollow ugc noopener\">documentation\u003C\u002Fa> for the full methodology and much more detailed evaluation results, or run your own evaluation on a project of your choice.\u003C\u002Fp>\n\u003Ch2>How Serena Works\u003C\u002Fh2>\n\u003Cp>Serena provides the necessary \u003Ca href=\"https:\u002F\u002Foraios.github.io\u002Fserena\u002F01-about\u002F035_tools.html\" rel=\"nofollow ugc noopener\">tools\u003C\u002Fa> for coding workflows, \nbut an LLM is required to do the actual work, orchestrating tool use.\u003C\u002Fp>\n\u003Cp>Serena can extend the functionality of your existing AI client via the \u003Cstrong>model context protocol (MCP)\u003C\u002Fstrong>.\nMost modern A\u003C\u002Fp>\n",1787530085978]