REQL helps coding agents fetch only the code context they need
REQL is a local repository for and . It is meant to help tools read a codebase through a compact, connected, source-grounded view instead of scanning the whole repository or depending only on whatever can fit into a prompt. It tracks relationships between files, symbols, imports, , tests, documentation, and other repository parts.
It uses a graph internally, but it is not a , graph framework, or graph visualization tool. The intended use is an embedded pipeline for repository indexing, s, querying, and context generation. It analyzes more than 30 s with Tree-sitter, with deeper support for Python, JavaScript, and TypeScript.
It includes , caching, deletion handling, watch mode, its own query language, local storage without an external , a CLI, a Python API, and an optional MCP server. Mandatory LLM calls are not required, so it could reduce token use and cost before an agent sends context to a model.
Key points
- REQL is a local for and .
- It maps relationships between files, symbols, imports, calls, tests, and documentation.
- It aims to avoid scanning an entire repository or stuffing too much code into a prompt.
- It supports more than 30 languages through Tree-sitter, with deeper support for Python, JavaScript, and TypeScript.
- It can run without mandatory LLM calls and offers a CLI, Python API, and optional MCP server.