ralfbecher/orionbelt-semantic-layer
API-first semantic layer engine with MCP server that compiles declarative YAML models into optimized, dialect-specific SQL across PostgreSQL, Snowflake, ClickHouse, Dremio, and Databricks.
What's novel
API-first semantic layer engine with MCP server that compiles declarative YAML models into optimized, dialect-specific SQL across PostgreSQL, Snowflake, ClickHouse, Dremio, and Databricks.
Code Analysis
6 files read · 2 roundsOrionBelt is a Python-based Semantic Layer Compiler that translates YAML-defined business logic (dimensions, measures, metrics) into optimized SQL queries for various database dialects via a FastAPI REST API.
Strengths
The project demonstrates excellent separation of concerns with a clear parser/compiler/API split. It handles complex semantic modeling (joins, aggregations, custom functions) and supports multiple database dialects with robust error reporting that suggests fixes to users.
Weaknesses
While the core logic is solid, the test coverage appears moderate; there's no evidence of extensive fuzzing or performance benchmarking tests. The novelty is limited as it solves a well-established problem (semantic layer generation) rather than introducing a fundamentally new paradigm.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
73
Contributors
1
Files
127
Active weeks
4
Repository
Language
Python
Stars
18
Forks
1
License
Apache-2.0