tree-sitter
24,948Incremental parser toolkit that powers structural code analysis and symbol discovery.
Maturity tier details
Computed tier: 2B(raw score 4.26). Effective tier: 2B.
Facets:documented_in=0evaluated_on=0integration_surfaces=0.5paper_citations=0recent_commit=2stars=1.76
Source of truth:pipeline/maturity/score.py
Description
Parser generator and incremental parsing library that produces concrete syntax trees for 100+ languages fast enough to reparse on every keystroke. Used as the structural-analysis substrate inside editors (Neovim, Helix, Zed), language servers, and code-intelligence pipelines - including duplicate-function detection, symbol extraction for RAG, and dead-code/unused-import linters. Ships as a C core plus language-specific grammar crates and bindings for Rust, Python, Node, Go, and Swift.
Why it's included
Occupies prevention x pre_generation x fully_autonomous for the dependency_blindness failure mode by providing the structural index that surfaces existing symbols, functions, and module exports before an agent reinvents them. Distinctive because it is the substrate many higher-level tools (LSPs, Sourcegraph SCIP emitters, grep-plus-AST linters) build on, which makes it the lowest-level entry point in the MECE grid for structural prevention - complementary to Semgrep (detection via custom rules) and sourcegraph-cody (retrieval grounding).
Security notes
MIT-licensed C library; grammars are community-maintained and compiled at build time. Running grammars on untrusted source is safe (pure parsing, no execution), but grammar-generation from .grammar.js uses node-gyp and should be pinned to audited grammar revisions to avoid supply-chain drift. No network access.
Addresses failure modes
- Dependency blindness — Reimplementing functionality that already exists in reachable code.