PRAXIS: Graph-Grounded Tacit Knowledge for Domain Code Generation
What it is
PRAXIS is a framework designed to help LLM agents generate domain-specific code by capturing 'tacit knowledge'—unwritten business rules and operational conventions. It works by simulating human development workflows to extract knowledge, organizing it into a code dependency graph, and surfacing it to the agent during code generation.
The problem it solves
LLM agents struggle with domain-specific code because tacit knowledge (e.g., business rules, conventions) is not explicitly documented in source code.
What you could use it for
- Generating code that adheres to complex, undocumented business logic
- Automating software engineering tasks in domains with strict interface contracts
- Improving the accuracy of LLM agents in large-scale enterprise codebases
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.