Serena Code Intelligence
Performs LSP-powered symbol navigation, type diagnostics, and AST-level code surgery for safe refactoring.
Serena Code Intelligence integrates Language Server Protocol (LSP) capabilities directly into your AI development workflow, enabling Claude to perform precise symbol navigation, inspect project-wide type diagnostics, and carry out targeted code surgery. Designed to solve the risks associated with raw string replacements and blind refactoring, Serena lets you track symbol declarations, find all implementations or referencing callers, safely rename or delete symbols across the entire repository, and persist complex blast-radius analysis across sessions using cross-session memory.
Key Features
01Cross-session memory system to store and recall architectural findings and blast-radius analysis
0223 GitHub stars
03Precise symbol navigation including declaration lookup, implementations, and cross-file references
04AST-level code surgery to safely replace function bodies, rename symbols, and delete unused definitions
05Real-time IDE-level diagnostics and type-error checks without needing to run entire test suites
06Seamless integration hooks for automated planning, TDD, execution, and code review workflows
Use Cases
01Executing large-scale architectural refactors and API signature renames with zero broken references
02Triaging type diagnostics and lint errors iteratively before triggering expensive CI or test suites
03Mapping caller impact and blast radius before modifying core interfaces or shared library utilities