tree-sitter

ast-grep rewrote Tree-sitter's C core in Rust, achieving up to 30% faster parsing and 22% faster end-to-end performance in ast-grep, at the cost of slightly higher memory usage.

Cameron Damera demonstrated Fred, a text editor written entirely in C, at Handmade Network Expo 2026. Its core features include an embedded TCC compiler, a live plugin system, dynamic Tree-sitter language support, an undo tree, and session recovery. All code is hand-written without AI assistance.

DeusData has open-sourced an MCP server called codebase-memory-mcp, which pre-parses code repositories into persistent knowledge graphs, providing AI coding assistants with sub-millisecond code structure query capabilities. It supports 158 languages, reduces token consumption by approximately 99% compared to traditional grep, and has extremely fast indexing speed.

Grove is a Rust-powered tool that uses tree-sitter to provide coding agents with structured, byte-precise, token-efficient access to codebases, supporting 27 languages via CLI and MCP server.

Kyde is a fast native Git client and code editor built in Rust using Zed's gpui framework, featuring GPU rendering, side-by-side diffs, tree-sitter syntax highlighting, and a hand-tuned dark theme.

codebase-memory-mcp is a tool written in pure C that parses the entire codebase into a knowledge graph, supports 158 programming languages, is compatible with 11 AI coding agent tools, greatly improving AI's understanding of project structure and reducing token consumption.

Emacs 31 is approaching with improvements like simpler tree-sitter setup and a built-in markdown-ts-mode. The article details features the author has been using from the development branch, highlighting reduced configuration overhead.

Emacs 31 introduces tree-sitter improvements that simplify grammar installation and includes a new built-in markdown-ts-mode. The author has been daily driving these changes from the development branch.

CodeGraph is a local-first tool that converts codebases into structured knowledge graphs. It uses Tree-sitter to precisely parse ASTs and supports over 20 languages. Through the MCP protocol, it directly feeds AI coding agents like Claude and Cursor, helping quickly analyze code impact scope while reducing token consumption and tool call frequency.

archex is a local-first, deterministic tool that builds token-budgeted code context bundles for AI agents, using a full retrieval pipeline (BM25F, local embeddings, cross-encoder reranker, dependency-graph expansion) on your hardware with no API keys or telemetry, outperforming alternatives in recall and efficiency.

Weave is an entity-level semantic merge driver for Git that resolves conflicts by parsing code with tree-sitter, merging functions and classes instead of lines. It supports 28 languages and offers additional coordination and MCP server features for multi-agent workflows.

An open-source project uses Tree-sitter to parse code into a graph structure and store it in local SQLite, providing a code map for AI coding agents, thereby reducing token consumption and costs. On average, it saves 57% tokens and reduces costs by 25%. Supports tools like Claude Code, Cursor, aider, etc.

git-courer is a full Git MCP server written in Go that uses tree-sitter for semantic code analysis, communicates via structured JSON, and supports 13 clients with local-first operation.

Spy-code is an open-source tool that builds a local codebase graph using tree-sitter, extracting functions, classes, and references to give AI coding agents a structured map of the codebase, currently supporting Rust, Python, TypeScript/JS, and Go.

The article explains why Tree-sitter is unsuitable for deep program analysis, highlighting how it discards critical tokens like operators and keywords. It advocates for using the Cubix framework as a more robust alternative for building semantic analysis and refactoring tools.

Octocode transforms code projects into navigable knowledge graphs for AI agents like Claude, Cursor, and Windsurf, using tree-sitter AST parsing and MCP integration to enable semantic search and dependency navigation.

Code-Graph-RAG is an open-source tool that parses multi-language codebases with Tree-sitter, builds a knowledge graph in Memgraph, and lets you query, edit, and optimize code using plain English. It supports monorepos with mixed languages under a unified graph schema.

code-review-graph is a Python tool that builds a structural map of code using Tree-sitter, tracks incremental changes, and provides precise context to AI coding assistants via MCP, significantly reducing token usage during code review tasks.

Codebase-memory-mcp is an ultra-fast code intelligence engine for AI coding agents that indexes entire repositories in milliseconds and answers structural queries in under 1ms using tree-sitter AST analysis and a persistent knowledge graph, with support for 158 languages and 14 MCP tools.