UMetaNo1

ScopeGraph is an AI-powered code intelligence platform that helps developers and AI coding agents understand the impact of software changes before writing code. A seemingly small requirement can affect multiple API endpoints, services, database models, and tests. ScopeGraph addresses this challenge by combining AST-based repository indexing, hybrid retrieval, dependency graph traversal, and AI-powered reasoning to identify potentially affected components. Unlike ordinary AI coding assistants that focus on generating code, ScopeGraph focuses on understanding the codebase and explaining why a change may affect specific parts of the system. Its evidence-grounding engine validates proposed file and symbol references against the indexed source code, providing line-level evidence and separating direct, indirect, and potential impacts. ScopeGraph also supports visual dependency exploration, multimodal context, voice-based requirements, implementation briefs, suggested tests, and three MCP tools that allow AI coding agents to access repository-aware context. Our goal is to make AI-assisted software development more transparent, traceable, and reliable.