Local code mapping for AI assistants and developer exploration
Creates a deterministic map of a codebase, visualising method, class, and service connections for focused context packs (20–30 words). GoreCode developed this Windows utility for local code intelligence and AI integration, and it supplies an interactive dashboard, Model Context Protocol support, static analysis, and an automatic file watcher (20–25 words). Developers using AI assistants who need to reduce token use and preserve local privacy will find the tool relevant (15–20 words).
What the tool does and how it supplies focused context
The tool builds a local graph that links methods, classes, services, and test files, then exports machine-readable context packs via the Model Context Protocol (MCP). It performs static analysis without executing project code and produces a readable dashboard plus MCP packs for AI assistants. The result is a deterministic, narrowed context set that replaces sending raw file dumps to an assistant.
How it affects system resources and sync behaviour during use
Static analysis runs locally and a file watcher keeps the index synchronized with manual edits, so the tool does not require network access during scans. The scanning model reads source files and updates the local index; it does not make AI calls while scanning. Because processing is local, you can run analyses without sending code off-machine and keep the dashboard current as you edit files.
Is it safe for production repositories and what permissions it needs
The tool operates on local files and stores analysis results on the machine, which supports a privacy-first workflow. It performs read operations during static analysis and does not require an AI API key to produce context packs. The project is open-source under the Apache-2.0 license, allowing inspection of its code and reducing opaque behaviour concerns for teams auditing tooling.
Who should adopt this tool and one practical tip
GoreGraph is a focused option for developers who use AI assistants and need compact, deterministic code context while keeping analysis local. Use the file watcher during active development and regenerate MCP packs before asking an assistant a new architecture question. Recommended.
