Getting Started with Code Analyzer

A step-by-step guide to indexing, searching, and reviewing your codebase. From zero to code intelligence in under 10 minutes.

Prerequisites

Requirement	Minimum	How to Check
Node.js	>= 20.0.0	`node --version`
pnpm	>= 9.0.0	`pnpm --version`
Git	Any recent version	`git --version`
Disk Space	~50 MB core + graph storage	Varies by codebase size

If you don't have pnpm installed:

npm install -g pnpm@latest
# Or via corepack (Node.js >= 16.13):
corepack enable && corepack prepare pnpm@latest --activate

Installation

Option 1 — Global Install (Recommended)

pnpm add -g @code-analyzer/cli

# Verify installation
code-analyzer --version

Expected output:

Code Analyzer v1.0.0
Node: v20.11.0 | Platform: linux x64

Option 2 — npx (No Install)

npx @code-analyzer/cli analyze --repo .

The first run downloads the package automatically. Subsequent runs are faster due to npx cache.

Option 3 — Docker

docker pull ghcr.io/agentix-e/code-analyzer:latest

# Run analysis on your project
docker run --rm -v $(pwd):/workspace ghcr.io/agentix-e/code-analyzer:latest \
  code-analyzer analyze --repo /workspace

# Start MCP server via Docker
docker run --rm -v $(pwd):/workspace -p 3100:3100 \
  ghcr.io/agentix-e/code-analyzer:latest \
  code-analyzer mcp --transport http --port 3100

Use Docker when you want a fully isolated environment or need to run Code Analyzer in CI/CD pipelines without installing Node.js.

First Steps

1. Initialize Your Project

Before your first analysis, initialize a configuration:

code-analyzer init

This creates a .code-analyzerrc file in your project root with sensible defaults. You'll be prompted to select:

Which languages to analyze (TypeScript, Python, Go, Java, Kotlin, C#, Rust)
Directories to exclude
Review severity preferences

Expected output:

✓ Created .code-analyzerrc
✓ Detected 3 languages: typescript, python, go
✓ Configuration saved with 8 options

2. Analyze Your Codebase

Run your first full analysis to build the knowledge graph:

code-analyzer analyze .

What happens during analysis:

File Discovery — Scans your project, respecting .gitignore and .code-analyzerignore
Parsing — Each source file is parsed by the appropriate language provider
Graph Building — A 19-phase DAG pipeline constructs a knowledge graph with 33 entity types and 44 relationship types
Indexing — Full-text and vector embeddings are generated for hybrid search

Expected output:

╔══════════════════════════════════════╗
║   Code Analyzer - Analysis Results   ║
╠══════════════════════════════════════╣
║ Files analyzed:        1,247         ║
║ Lines of code:         87,342        ║
║ Nodes created:         4,521         ║
║ Relationships created:  18,330        ║
║ Graph size:            12.4 MB       ║
║ Analysis time:         8.3s          ║
╚══════════════════════════════════════╝

Languages detected: typescript (847 files), python (312 files), go (88 files)

3. Search Your Code

Now that your codebase is indexed, search it:

# Keyword search
code-analyzer search "authentication"

# Semantic search (what does this code do?)
code-analyzer search "how does the login flow work" --semantic

# Search with filters
code-analyzer search "handler" --language typescript --type Function

# Cypher graph query
code-analyzer search --cypher "MATCH (f:Function) WHERE f.name CONTAINS 'auth' RETURN f.name, f.file"

Expected output (keyword search):

Search: "authentication" (BM25, top 20 results)

1. auth/login.ts:42  —  authenticateUser()      [score: 0.892]
2. auth/middleware.ts:18  —  authMiddleware()    [score: 0.845]
3. services/token.ts:67  —  refreshAuthToken()    [score: 0.801]
4. types/auth.ts:5  —  AuthConfig interface       [score: 0.763]
...

Found 47 results in 0.12s

4. Review Your Code

Get automated review feedback on your code:

# Review a single file
code-analyzer review src/auth/login.ts

# Review staged changes (before committing)
code-analyzer review src/ --diff

# Review entire directory against standards
code-analyzer review src/ --standard typescript-best-practices

Expected output:

Review: src/auth/login.ts
═══════════════════════════════
[CRITICAL] Line 42: Hardcoded secret - API key appears to be embedded in source
  → Move to environment variable or secrets manager

[HIGH] Line 67: Missing error handling - async function lacks try/catch
  → Wrap database call in try/catch with appropriate error response

[MEDIUM] Line 89: Function length exceeds threshold (52 lines)
  → Consider refactoring into smaller functions

[LOW] Line 12: Unused import 'crypto' detected
  → Remove unused import

Summary: 1 critical, 1 high, 1 medium, 1 low — 4 issues total

Setting Up MCP for AI Agents

Code Analyzer exposes 45 tools via the Model Context Protocol (MCP), turning your AI coding agent into a code intelligence powerhouse.

Auto-Detect and Configure

The easiest way to set up MCP is with the agent detection command:

code-analyzer agent detect

This scans your environment for supported AI agents and shows what's available:

Agent Detection Results
═════════════════════════
✓ Claude Desktop detected — ~/Library/Application Support/Claude/claude_desktop_config.json
✓ Cursor detected — .cursor/mcp.json
✓ VS Code detected — Code Analyzer extension installed
✓ Windsurf detected — ~/.windsurf/mcp.json

Run 'code-analyzer agent configure' to set up all detected agents.

Then configure all detected agents at once:

code-analyzer agent configure

Expected output:

✓ Configured Claude Desktop (38 tools)
✓ Configured Cursor (28 tools, analysis profile)
✓ Configured Windsurf (28 tools, analysis profile)

Restart your AI agents to begin using Code Analyzer tools.

Manual Configuration

If auto-detection doesn't work, configure manually. For Claude Desktop, add to claude_desktop_config.json:

{
  "mcpServers": {
    "code-analyzer": {
      "command": "npx",
      "args": ["-y", "@code-analyzer/mcp"],
      "env": {
        "CODE_ANALYZER_PROJECT_DIR": "/absolute/path/to/your/project"
      }
    }
  }
}

For Cursor, create .cursor/mcp.json in your project root:

{
  "mcpServers": {
    "code-analyzer": {
      "command": "npx",
      "args": ["-y", "@code-analyzer/mcp"],
      "env": {
        "CODE_ANALYZER_PROJECT_DIR": "${workspaceFolder}"
      }
    }
  }
}

After restarting your AI agent, you'll see a hammer icon in the chat interface, confirming the 45 MCP tools are available. See the MCP Tool Reference for the complete listing.

VS Code Extension Setup

Installation

Open VS Code
Press Ctrl+Shift+X (or Cmd+Shift+X on macOS)
Search for "Code Analyzer"
Click Install

Alternatively, install from the VS Code Marketplace.

Verify Installation

After installation, you should see:

Activity Bar Icon — The Code Analyzer icon (magnifying glass over brackets) appears in the activity bar
Status Bar Indicator — Shows "CA: Indexed" when a project is analyzed
Output Panel — View → Output → "Code Analyzer" shows extension logs

Key Features

Feature	How to Access
Knowledge Graph Sidebar	Click the Code Analyzer icon in the activity bar
Copilot Chat Integration	Type `@code-analyzer` in Copilot Chat (requires GitHub Copilot)
Inline Review Comments	Hover over code to see AI review suggestions
Impact Analysis	Right-click a function → "Code Analyzer: Analyze Impact"
Command Palette	`Ctrl+Shift+P` → search "Code Analyzer"

Copilot Chat Commands

@code-analyzer /review    — Review the current file
@code-analyzer /explain   — Explain selected code
@code-analyzer /impact    — Analyze impact of current function
@code-analyzer /find      — Search for symbols
@code-analyzer /deps      — Show dependencies
@code-analyzer /refactor  — Suggest refactoring
@code-analyzer /test      — Generate tests for current file
@code-analyzer /coverage  — Show test coverage gaps
@code-analyzer /standards — Check against standards

Extension Settings

Configure via Ctrl+, → search "Code Analyzer":

Setting	Default	Description
`codeAnalyzer.indexOnOpen`	`true`	Auto-index workspace when opened
`codeAnalyzer.languages`	`["typescript","javascript"]`	Languages to analyze
`codeAnalyzer.autoReview`	`false`	Automatically review on file save
`codeAnalyzer.ignorePatterns`	`["node_modules","dist"]`	Patterns to skip

Troubleshooting Common First-Time Issues

"command not found: code-analyzer"

The global install path isn't in your $PATH. Run:

pnpm setup
source ~/.bashrc  # or ~/.zshrc

Or use npx directly: npx @code-analyzer/cli analyze .

"No files found to analyze"

Check that:

You're in a directory with supported source files (.ts, .py, .go, .java, etc.)
Your files aren't excluded by .gitignore patterns
You've specified the right language: code-analyzer analyze . --languages typescript

"Analysis is slow" on Large Projects

Limit languages: code-analyzer analyze . --languages typescript

Exclude generated files in .code-analyzerrc:

{ "excludePatterns": ["**/generated/**", "**/*.generated.*"] }

Set CODE_ANALYZER_PARSE_WORKERS=8 for more parallel workers

"MCP server won't start"

Check for port conflicts: lsof -i :3100
Verify Node.js version: node --version (must be >= 20)
Run directly to see errors: npx @code-analyzer/mcp --transport http --port 3100

"VS Code extension shows nothing"

Check the extension is activated: View → Output → select "Code Analyzer"
Reload VS Code: Ctrl+Shift+P → "Developer: Reload Window"
Verify Node.js >= 20 is installed and on $PATH

Next Steps

Resource	Description
Configuration Reference	All config options and environment variables
MCP Tool Reference	Complete 45-tool reference for AI agents
Scenario Guides	Task-based workflows (PR review, CI/CD, monorepo)
Troubleshooting	Detailed solutions for common issues
Architecture	Deep dive into the system design
Language Support	Supported languages and feature matrix

Navigation Menu

Uh oh!

Getting Started with Code Analyzer

Prerequisites

Installation

Option 1 — Global Install (Recommended)

Option 2 — npx (No Install)

Option 3 — Docker

First Steps

1. Initialize Your Project

2. Analyze Your Codebase

3. Search Your Code

4. Review Your Code

Setting Up MCP for AI Agents

Auto-Detect and Configure

Manual Configuration

VS Code Extension Setup

Installation

Verify Installation

Key Features

Copilot Chat Commands

Extension Settings

Troubleshooting Common First-Time Issues

"command not found: code-analyzer"

"No files found to analyze"

"Analysis is slow" on Large Projects

"MCP server won't start"

"VS Code extension shows nothing"

Next Steps

Uh oh!

FilesExpand file tree

getting-started.md

Latest commit

History

getting-started.md

File metadata and controls

Getting Started with Code Analyzer

Prerequisites

Installation

Option 1 — Global Install (Recommended)

Option 2 — npx (No Install)

Option 3 — Docker

First Steps

1. Initialize Your Project

2. Analyze Your Codebase

3. Search Your Code

4. Review Your Code

Setting Up MCP for AI Agents

Auto-Detect and Configure

Manual Configuration

VS Code Extension Setup

Installation

Verify Installation

Key Features

Copilot Chat Commands

Extension Settings

Troubleshooting Common First-Time Issues

"command not found: code-analyzer"

"No files found to analyze"

"Analysis is slow" on Large Projects

"MCP server won't start"

"VS Code extension shows nothing"

Next Steps