QE Code Intelligence

Purpose

Guide the use of v3's code intelligence capabilities including knowledge graph construction, semantic code search, dependency mapping, and context-aware code understanding with significant token reduction.

Activation

• When understanding unfamiliar code
• When searching for code semantically
• When analyzing dependencies
• When building code knowledge graphs
• When reducing context for AI operations

Quick Start

bash
1# Index codebase into knowledge graph
2aqe code index src/ --incremental
3
4# Semantic code search
5aqe code search "authentication middleware"
6
7# Analyze change impact
8aqe code impact src/services/UserService.ts --depth 3
9
10# Map dependencies
11aqe code deps src/
12
13# Analyze complexity and find hotspots
14aqe code complexity src/

Agent Workflow

typescript
1// Build knowledge graph
2Task("Index codebase", `
3  Build knowledge graph for the project:
4  - Parse all TypeScript files in src/
5  - Extract entities (classes, functions, types)
6  - Map relationships (imports, calls, inheritance)
7  - Generate embeddings for semantic search
8  Store in AgentDB vector database.
9`, "qe-kg-builder")
10
11// Semantic search
12Task("Find relevant code", `
13  Search for code related to "user authentication flow":
14  - Use semantic similarity (not just keyword)
15  - Include related functions and types
16  - Rank by relevance score
17  - Return with minimal context (80% token reduction)
18`, "qe-code-intelligence")

Knowledge Graph Operations

1. Codebase Indexing

typescript
1await knowledgeGraph.index({
2  source: 'src/**/*.ts',
3  extraction: {
4    entities: ['class', 'function', 'interface', 'type', 'variable'],
5    relationships: ['imports', 'calls', 'extends', 'implements', 'uses'],
6    metadata: ['jsdoc', 'complexity', 'lines']
7  },
8  embeddings: {
9    model: 'code-embedding',
10    dimensions: 384,
11    normalize: true
12  },
13  incremental: true  // Only index changed files
14});

2. Semantic Search

typescript
1await semanticSearcher.search({
2  query: 'payment processing with stripe',
3  options: {
4    similarity: 'cosine',
5    threshold: 0.7,
6    limit: 20,
7    includeContext: true
8  },
9  filters: {
10    fileTypes: ['.ts', '.tsx'],
11    excludePaths: ['node_modules', 'dist']
12  }
13});

3. Dependency Analysis

typescript
1await dependencyMapper.analyze({
2  entry: 'src/services/OrderService.ts',
3  depth: 3,
4  direction: 'both',  // imports and importedBy
5  output: {
6    graph: true,
7    metrics: {
8      afferentCoupling: true,
9      efferentCoupling: true,
10      instability: true
11    }
12  }
13});

Token Reduction Strategy

typescript
1// Get context with 80% token reduction
2const context = await codeIntelligence.getOptimizedContext({
3  query: 'implement user registration',
4  budget: 4000,  // max tokens
5  strategy: {
6    relevanceRanking: true,
7    summarization: true,
8    codeCompression: true,
9    deduplication: true
10  },
11  include: {
12    signatures: true,
13    implementations: 'relevant-only',
14    comments: 'essential',
15    examples: 'top-3'
16  }
17});

Knowledge Graph Schema

typescript
1interface KnowledgeGraph {
2  entities: {
3    id: string;
4    type: 'class' | 'function' | 'interface' | 'type' | 'file';
5    name: string;
6    file: string;
7    line: number;
8    embedding: number[];
9    metadata: Record<string, any>;
10  }[];
11  relationships: {
12    source: string;
13    target: string;
14    type: 'imports' | 'calls' | 'extends' | 'implements' | 'uses';
15    weight: number;
16  }[];
17  indexes: {
18    byName: Map<string, string[]>;
19    byFile: Map<string, string[]>;
20    byType: Map<string, string[]>;
21  };
22}

Search Results

typescript
1interface SearchResult {
2  entity: {
3    name: string;
4    type: string;
5    file: string;
6    line: number;
7  };
8  relevance: number;
9  snippet: string;
10  context: {
11    before: string[];
12    after: string[];
13    related: string[];
14  };
15  explanation: string;
16}

CLI Examples

bash
1# Full reindex
2aqe code index src/
3
4# Incremental index (changed files only)
5aqe code index src/ --incremental
6
7# Index only files changed since a git ref
8aqe code index . --git-since HEAD~5
9
10# Semantic code search
11aqe code search "database connection"
12
13# Change impact analysis
14aqe code impact src/services/UserService.ts
15
16# Dependency mapping
17aqe code deps src/ --depth 5
18
19# Complexity metrics and hotspots
20aqe code complexity src/ --format json

Gotchas

• WARNING: code-intelligence domain has 18% success rate — prefer direct grep/glob over agent-based code search for simple queries
• Knowledge graph construction fails on repos >50K LOC — scope to specific modules
• Semantic search returns irrelevant results without domain-specific embeddings — always verify search results manually
• Agent claims "80% token reduction" but may skip critical context — verify key files are included in results
• Fleet must be initialized before using: run aqe health to diagnose, or aqe init to re-initialize if you get initialization errors

Coordination

Primary Agents: qe-kg-builder, qe-dependency-mapper, qe-impact-analyzer, qe-code-complexity Coordinator: qe-code-intelligence Related Skills: qe-test-generation, qe-defect-intelligence