codebase-context

Overview

Codebase-context generates a map of your codebase to help AI agents understand your architecture, coding conventions, and patterns before they start searching. It provides bounded conventions mapping, pattern detection, semantic search with quality indicators, and team memory functionality. The server supports both stdio and HTTP modes, works with multiple AI clients, and offers multi-project support with different languages.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• search_codebase

  Hybrid search + decision card when intent='edit'

  Try:Search for auth middleware implementationTry:Find examples of data validation patterns
• get_team_patterns

  Pattern frequencies, golden files, conflict detection

  Try:What patterns are being used for state management?Try:Show me the most commonly used testing frameworks
• get_symbol_references

  Concrete references to a symbol (count + snippets)

  Try:Where is the AuthService class being used?Try:Find all references to the calculateTotal function
• remember

  Record a convention, decision, gotcha, or failure

  Try:Remember that we use dependency injection for all servicesTry:Record gotcha: avoid using setTimeout for timers
• get_memory

  Query team memory with confidence decay scoring

  Try:What are the team conventions for API authentication?Try:Get team decisions about error handling patterns
• get_codebase_metadata

  Project structure, frameworks, dependencies

  Try:What frameworks are being used in this project?Try:Show me the project structure and main directories
• get_style_guide

  Style guide rules for the current project

  Try:What's the code style for naming functions?Try:Show me the linting rules for this project
• detect_circular_dependencies

  Import cycles between files

  Try:Check for circular dependencies in the auth moduleTry:Detect any import cycles between services
• refresh_index

  Full or incremental re-index + git memory extraction

  Try:Refresh the codebase index to include recent changesTry:Do an incremental update of the codebase index
• get_indexing_status

  Progress and stats for the current index

  Try:Check the status of the codebase indexingTry:Show me statistics about the last index update

Comparable tools

Installation

{
  "mcpServers": {
    "codebase-context": {
      "command": "npx",
      "args": ["-y", "codebase-context"]
    }
  }
}

FAQ

How does codebase-context differ from code search tools?

It starts with a bounded conventions map showing architecture layers and patterns before searching, rather than dumping raw search results. It also provides pattern trend analysis and quality indicators.

What languages are supported?

Full symbol extraction for 10 languages (TS, JS, Python, Java, Kotlin, C, C++, C#, Go, Rust) and indexing coverage for 30+ languages including PHP, Ruby, Swift, and Shell.

Compare codebase-context with