kythe/kythe

Health v2 · maintenance only

Flags: no_readme

Adoption not part of the score

2147 stars · 273 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Kythe is a pluggable, mostly language-agnostic ecosystem for building tools that work with code, providing a standard schema and interchange format for semantic code information. It includes indexers for C++ and Java, compilation extractors, a verifier, and a cross-reference service.

Use cases

• build a code search and cross-reference service for a large codebase
• index C++ and Java sources to power go-to-definition and find-references in editors
• share semantic code analysis data between build systems, compilers, and static analysis tools
• write a custom language indexer against a standard schema
• verify that an indexer produces correct code graphs
• extract compilation information from javac, Maven, cmake, Go, and Bazel builds

When to choose

• you need language-agnostic code intelligence infrastructure across a large, multi-language codebase
• you want to build custom code navigation, cross-reference, or static analysis tooling
• you need a standard interchange format (Kythe schema, kzip) for code analysis data

When to avoid

• you just need simple IDE navigation for a single language - a language server is simpler
• you need a turnkey hosted code search product with minimal setup
• your project is small and does not justify running an indexing pipeline

Facets

framework · maturity active

search-engine parser developer-tools documentation developer-tools parsers programming-languages apis go cpp jvm cross-platform code-indexing cross-references static-analysis code-navigation language-agnostic code-intelligence linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="kythe/kythe")