G
Graphify
graphify.homes
Sign in
GitHub Copilot context

GitHub Copilot codebase context beyond file search

GitHub Copilot codebase context improves when repository instructions explain how the team works and a project graph explains how the code is connected. Graphify adds graph.html, graph.json, and GRAPH_REPORT.md alongside Copilot's source search, so the assistant can follow dependencies, callers, and architecture communities instead of rediscovering structure from filenames.

Published Updated

Graphify outputs
graph.html
graph.json
GRAPH_REPORT.md

Graphify stores inspectable artifacts so codebase knowledge can be reviewed by humans and reused by AI coding assistants.

A Copilot-style editor surface connected to a persistent codebase knowledge graph

Instructions

Repository rules

Use Copilot repository instructions for commands, conventions, and project constraints.

Structure

Graph artifacts

Use Graphify for dependencies, call paths, communities, and project documentation relationships.

Tools

MCP

Expose graph queries as repeatable tools when the Copilot surface supports MCP.

What is GitHub Copilot codebase context?

GitHub Copilot codebase context is the combination of source search, repository instructions, open files, tool results, and external services available while Copilot works. Better context helps Copilot find the right implementation and respect how the repository is maintained.

  • Source files provide exact implementation behavior.
  • Repository custom instructions provide durable team guidance.
  • A code graph provides relationships that file search may not surface.

How do Copilot repository instructions help?

Repository custom instructions describe the commands, conventions, architecture rules, and review expectations Copilot should follow. They work best as concise operating guidance, not as a hand-maintained inventory of every file.

  • Document build, test, lint, and validation commands.
  • Explain protected paths and required review workflows.
  • Tell Copilot when to query the graph before a broad search.

What does a code graph add to Copilot?

A graph adds explicit relationships between files, symbols, imports, calls, documents, and rationale. Copilot can ask where a change may travel, which modules own a responsibility, and which documentation explains a dependency.

  • Trace incoming and outgoing code relationships.
  • Expose subsystem communities and high-connectivity nodes.
  • Connect implementation to design docs and project context.
  • Use confidence labels to separate parsed relationships from inferred ones.

How do you connect Graphify to Copilot?

Use the installer for the available Copilot surface, build a project graph, and add instructions that tell Copilot when to use the artifacts. The current Graphify CLI exposes both copilot and VS Code installation workflows.

  • Run graphify copilot install for the supported Copilot integration.
  • Run graphify vscode install when configuring Copilot Chat in VS Code.
  • Refresh the graph after large structural changes.

When should Copilot use the graph instead of grep?

Use the graph for architecture, dependency, call-path, onboarding, and blast-radius questions. Use grep for exact strings, configuration values, recent edits, and code that the graph may not index.

  • Graph first for relationships between known symbols.
  • Grep first for exact text or a narrow recent change.
  • Open source files before making any edit.

How should teams verify Copilot answers?

Require Copilot to cite source files and tests before accepting an architectural claim. Graph paths are review candidates; they do not prove runtime behavior or replace compilation and test execution.

  • Verify inferred graph edges in source.
  • Run the tests selected from affected paths.
  • Confirm the graph reflects the current commit.
Working commands

How do the setup commands fit together?

Install the Copilot integration

Shell
graphify copilot install
FAQ

Frequently asked questions

Does Graphify replace Copilot source search?

No. It adds structural context and navigation. Source search remains useful for exact text and recent changes.

Can Copilot use graph.json directly?

Copilot can read project artifacts when its surface allows file access. MCP provides a more structured interface for graph tools.

Which repository file should hold the graph policy?

Use the repository instruction mechanism supported by the specific Copilot surface and reference graph.json instead of copying the whole report.

Does a graph make Copilot answers automatic?

No. It improves context and navigation; the team still verifies source, tests, security, and product intent.

Key takeaways

What should you remember?

  • Use repository instructions for team policy and Graphify for project relationships.
  • Copilot should query the graph for architecture and impact, then verify source and tests.
  • Choose the copilot or vscode installer based on the actual surface being used.
Sources

Where can you verify the details?

GitHub Copilot repository instructionsOfficial guidance for repository-level Copilot instructions.Graphify GitHub repositoryGraph artifacts, assistant installers, queries, and confidence labels.Graphify architecture documentationHow Graphify combines extracted structure and semantic context.

Related Graphify guides

Continue through the Graphify topic cluster with pages that explain assistant context, architecture review, documentation, and graph-based codebase understanding.

Codex codebase context without rereading the repository

Give Codex persistent codebase context with graph artifacts, concise AGENTS.md rules, and MCP tools so repository questions use structure, not repeated scans.

Cursor codebase context beyond indexing

Cursor indexing finds similar code. Graphify adds structural context: dependencies, subsystem clusters, and rationale your editor cannot infer from embeddings alone.

MCP codebase knowledge graph for AI coding assistants

Use an MCP codebase knowledge graph to give Codex, Claude, and other agents structured queries for dependencies, architecture, and project context.

Turn this concept into a real project graph

Start from the use cases page or upload a project ZIP to build a graph that your team can inspect and reuse.

See use casesBuild a graph