Module ai_contract_diff

Expand description

ai_contract_diff lives in mockforge_intelligence::ai_contract_diff (Issue #562 phase 4). Re-exported here so existing crate::ai_contract_diff::* call sites inside core (ai_studio / contract_diff_handler, request_capture) and external mockforge_core::ai_contract_diff::* consumers (mockforge-http, mockforge-cli) keep compiling unchanged. Pillars: [Contracts][AI]

AI-powered contract diff analysis

This module provides intelligent contract diff analysis that compares front-end requests against backend API contract specifications, detects mismatches, and generates AI-powered recommendations and correction proposals.

§Features

Structural Diff Analysis: Detects mismatches between requests and contract specs
AI-Powered Recommendations: Uses LLM to generate contextual recommendations
Correction Proposals: Generates JSON Patch files for schema corrections
Confidence Scoring: Provides confidence scores for all suggestions

§Example Usage

use mockforge_core::ai_contract_diff::{
    ContractDiffAnalyzer, ContractDiffConfig, CapturedRequest,
};
use mockforge_core::openapi::OpenApiSpec;

async fn example() -> mockforge_core::Result<()> {
    // Load contract specification
    let spec = OpenApiSpec::from_file("api.yaml").await?;

// Configure contract diff
let config = ContractDiffConfig {
    enabled: true,
    llm_provider: "openai".to_string(),
    llm_model: "gpt-4".to_string(),
    confidence_threshold: 0.5,
    ..Default::default()
};

// Create analyzer
let analyzer = ContractDiffAnalyzer::new(config)?;

// Capture a request
let request = CapturedRequest::new("POST", "/api/users", "browser_extension")
    .with_body(serde_json::json!({"name": "Alice", "email": "alice@example.com"}));

// Analyze request against contract
let result = analyzer.analyze(request, &spec).await?;

// Check results
if !result.matches {
    println!("Found {} mismatches", result.mismatches.len());
    for mismatch in &result.mismatches {
        println!("  - {}: {}", mismatch.path, mismatch.description);
    }

    // Generate recommendations
    for recommendation in &result.recommendations {
        println!("  Recommendation: {}", recommendation.recommendation);
    }

    // Generate correction proposals
    for correction in &result.corrections {
        println!("  Correction: {}", correction.description);
    }
    }
    Ok(())
}

Modules§

confidence_scorer: Confidence scoring system for AI contract diff recommendations
correction_proposer: Schema correction proposal generator
diff_analyzer: Core diff analysis engine for contract comparison
recommendation_engine: AI-powered recommendation engine for contract diff analysis
semantic_analyzer: Semantic drift analysis for contract diffs
types: Re-exported from mockforge_foundation::contract_diff_types (Phase 6 / A5).

Structs§

CapturedRequest: Captured request for contract analysis
ConfidenceScorer: Confidence scorer for contract diff recommendations
ContractDiffAnalyzer: Main contract diff analyzer that orchestrates all components
ContractDiffConfig: Configuration for contract diff analysis
ContractDiffResult: Result of contract diff analysis between a request and a contract specification
CorrectionProposal: Correction proposal with patch operations
CorrectionProposer: Correction proposal generator
DiffAnalyzer: Contract diff analyzer
DiffMetadata: Metadata about the diff analysis
Mismatch: A detected mismatch between request and contract
Recommendation: AI-generated recommendation for fixing a mismatch
RecommendationEngine: AI-powered recommendation engine
RequestContext: Context for recommendation generation
ScoringContext: Context for confidence scoring
SemanticAnalyzer: Semantic analyzer for detecting meaning changes
SemanticDriftResult: Semantic drift analysis result

Enums§

ConfidenceLevel: Confidence level categories for visual indicators
MismatchSeverity: Severity levels for mismatches
MismatchType: Types of mismatches that can be detected
PatchOperation: JSON Patch operation types (RFC 6902)
SemanticChangeType: Type of semantic change