zero-script-qa（零脚本质量保证）

Zero Script QA Expert Knowledge

Overview

Zero Script QA is a methodology that verifies features through structured logs and real-time monitoring without writing test scripts.

Traditional: Write test code → Execute → Check results → Maintain
Zero Script: Build log infrastructure → Manual UX test → AI log analysis → Auto issue detection

Core Principles

1. Log Everything

• All API calls (including 200 OK)
• All errors
• All important business events
• Entire flow trackable via Request ID

2. Structured JSON Logs

• Parseable JSON format
• Consistent fields (timestamp, level, request_id, message, data)
• Different log levels per environment

3. Real-time Monitoring

• Docker log streaming
• Claude Code analyzes in real-time
• Immediate issue detection and documentation

Logging Architecture

JSON Log Format Standard

{
  "timestamp": "2026-01-08T10:30:00.000Z",
  "level": "INFO",
  "service": "api",
  "request_id": "req_abc123",
  "message": "API Request completed",
  "data": {
    "method": "POST",
    "path": "/api/users",
    "status": 200,
    "duration_ms": 45
  }
}

Required Log Fields

Log Level Policy

Request ID Propagation

Concept

Client → API Gateway → Backend → Database
   ↓         ↓           ↓          ↓
req_abc   req_abc     req_abc    req_abc

Trackable with same Request ID across all layers

Implementation Patterns

1. Request ID Generation (Entry Point)

// middleware.ts
import { v4 as uuidv4 } from 'uuid';

export function generateRequestId(): string {
  return `req_${uuidv4().slice(0, 8)}`;
}

// Propagate via header
headers['X-Request-ID'] = requestId;

2. Request ID Extraction and Propagation

// API client
const requestId = headers['X-Request-ID'] || generateRequestId();

// Include in all logs
logger.info('Processing request', { request_id: requestId });

// Include in header when calling downstream services
await fetch(url, {
  headers: { 'X-Request-ID': requestId }
});

Backend Logging (FastAPI)

Logging Middleware

# middleware/logging.py
import logging
import time
import uuid
import json
from fastapi import Request

class JsonFormatter(logging.Formatter):
    def format(self, record):
        log_record = {
            "timestamp": self.formatTime(record),
            "level": record.levelname,
            "service": "api",
            "request_id": getattr(record, 'request_id', 'N/A'),
            "message": record.getMessage(),
        }
        if hasattr(record, 'data'):
            log_record["data"] = record.data
        return json.dumps(log_record)

class LoggingMiddleware:
    async def __call__(self, request: Request, call_next):
        request_id = request.headers.get('X-Request-ID', f'req_{uuid.uuid4().hex[:8]}')
        request.state.request_id = request_id

        start_time = time.time()

        # Request logging
        logger.info(
            f"Request started",
            extra={
                'request_id': request_id,
                'data': {
                    'method': request.method,
                    'path': request.url.path,
                    'query': str(request.query_params)
                }
            }
        )

        response = await call_next(request)

        duration = (time.time() - start_time) * 1000

        # Response logging (including 200 OK!)
        logger.info(
            f"Request completed",
            extra={
                'request_id': request_id,
                'data': {
                    'status': response.status_code,
                    'duration_ms': round(duration, 2)
                }
            }
        )

        response.headers['X-Request-ID'] = request_id
        return response

Business Logic Logging

# services/user_service.py
def create_user(data: dict, request_id: str):
    logger.info("Creating user", extra={
        'request_id': request_id,
        'data': {'email': data['email']}
    })

    # Business logic
    user = User(**data)
    db.add(user)
    db.commit()

    logger.info("User created", extra={
        'request_id': request_id,
        'data': {'user_id': user.id}
    })

    return user

Frontend Logging (Next.js)

Logger Module

// lib/logger.ts
type LogLevel = 'DEBUG' | 'INFO' | 'WARNING' | 'ERROR';

interface LogData {
  request_id?: string;
  [key: string]: any;
}

const LOG_LEVELS: Record<LogLevel, number> = {
  DEBUG: 0,
  INFO: 1,
  WARNING: 2,
  ERROR: 3,
};

const MIN_LEVEL = process.env.NODE_ENV === 'production' ? 'INFO' : 'DEBUG';

function log(level: LogLevel, message: string, data?: LogData) {
  if (LOG_LEVELS[level] < LOG_LEVELS[MIN_LEVEL]) return;

  const logEntry = {
    timestamp: new Date().toISOString(),
    level,
    service: 'web',
    request_id: data?.request_id || 'N/A',
    message,
    data: data ? { ...data, request_id: undefined } : undefined,
  };

  console.log(JSON.stringify(logEntry));
}

export const logger = {
  debug: (msg: string, data?: LogData) => log('DEBUG', msg, data),
  info: (msg: string, data?: LogData) => log('INFO', msg, data),
  warning: (msg: string, data?: LogData) => log('WARNING', msg, data),
  error: (msg: string, data?: LogData) => log('ERROR', msg, data),
};

API Client Integration

// lib/api-client.ts
import { logger } from './logger';
import { v4 as uuidv4 } from 'uuid';

export async function apiClient<T>(
  endpoint: string,
  options: RequestInit = {}
): Promise<T> {
  const requestId = `req_${uuidv4().slice(0, 8)}`;
  const startTime = Date.now();

  logger.info('API Request started', {
    request_id: requestId,
    method: options.method || 'GET',
    endpoint,
  });

  try {
    const response = await fetch(`/api${endpoint}`, {
      ...options,
      headers: {
        'Content-Type': 'application/json',
        'X-Request-ID': requestId,
        ...options.headers,
      },
    });

    const duration = Date.now() - startTime;
    const data = await response.json();

    // Log 200 OK too!
    logger.info('API Request completed', {
      request_id: requestId,
      status: response.status,
      duration_ms: duration,
    });

    if (!response.ok) {
      logger.error('API Request failed', {
        request_id: requestId,
        status: response.status,
        error: data.error,
      });
      throw new ApiError(data.error);
    }

    return data;
  } catch (error) {
    logger.error('API Request error', {
      request_id: requestId,
      error: error instanceof Error ? error.message : 'Unknown error',
    });
    throw error;
  }
}

Nginx JSON Logging

nginx.conf Configuration

http {
    log_format json_combined escape=json '{'
        '"timestamp":"$time_iso8601",'
        '"level":"INFO",'
        '"service":"nginx",'
        '"request_id":"$http_x_request_id",'
        '"message":"HTTP Request",'
        '"data":{'
            '"remote_addr":"$remote_addr",'
            '"method":"$request_method",'
            '"uri":"$request_uri",'
            '"status":$status,'
            '"body_bytes_sent":$body_bytes_sent,'
            '"request_time":$request_time,'
            '"upstream_response_time":"$upstream_response_time",'
            '"http_referer":"$http_referer",'
            '"http_user_agent":"$http_user_agent"'
        '}'
    '}';

    access_log /var/log/nginx/access.log json_combined;
}

Docker-Based QA Workflow

docker-compose.yml Configuration

version: '3.8'
services:
  api:
    build: ./backend
    environment:
      - LOG_LEVEL=DEBUG
      - LOG_FORMAT=json
    logging:
      driver: json-file
      options:
        max-size: "10m"
        max-file: "3"

  web:
    build: ./frontend
    environment:
      - NODE_ENV=development
    depends_on:
      - api

  nginx:
    image: nginx:alpine
    volumes:
      - ./nginx/nginx.conf:/etc/nginx/nginx.conf
    ports:
      - "80:80"
    depends_on:
      - api
      - web

Real-time Log Monitoring

# Stream all service logs
docker compose logs -f

# Specific service only
docker compose logs -f api

# Filter errors only
docker compose logs -f | grep '"level":"ERROR"'

# Track specific Request ID
docker compose logs -f | grep 'req_abc123'

QA Automation Workflow

1. Start Environment

# Start development environment
docker compose up -d

# Start log monitoring (Claude Code monitors)
docker compose logs -f

2. Manual UX Testing

User tests actual features in browser:
1. Sign up attempt
2. Login attempt
3. Use core features
4. Test edge cases

3. Claude Code Log Analysis

Claude Code in real-time:
1. Monitor log stream
2. Detect error patterns
3. Detect abnormal response times
4. Track entire flow via Request ID
5. Auto-document issues

4. Issue Documentation

# QA Issue Report

## Issues Found

### ISSUE-001: Insufficient error handling on login failure
- **Request ID**: req_abc123
- **Severity**: Medium
- **Reproduction path**: Login → Wrong password
- **Log**:

{"level":"ERROR","message":"Login failed","data":{"error":"Invalid credentials"}}

- **Problem**: Error message not user-friendly
- **Recommended fix**: Add error code to message mapping

Issue Detection Patterns

1. Error Detection

{"level":"ERROR","message":"..."}

→ Report immediately

2. Slow Response Detection

{"data":{"duration_ms":3000}}

→ Warning when exceeding 1000ms

3. Consecutive Failure Detection

3+ consecutive failures on same endpoint

→ Report potential system issue

4. Abnormal Status Codes

{"data":{"status":500}}

→ Report 5xx errors immediately

Phase Integration

Iterative Test Cycle Pattern

Based on bkamp.ai notification feature development:

Example: 8-Cycle Test Process

Cycle Workflow

┌─────────────────────────────────────────────────────────────┐
│                   Iterative Test Cycle                        │
├─────────────────────────────────────────────────────────────┤
│                                                             │
│  Cycle N:                                                   │
│  1. Run test script (E2E or manual)                         │
│  2. Claude monitors logs in real-time                       │
│  3. Record pass/fail results                                │
│  4. Claude identifies root cause of failures                │
│  5. Fix code immediately (hot reload)                       │
│  6. Document: Cycle N → Bug → Fix                           │
│                                                             │
│  Repeat until acceptable pass rate (>85%)                   │
│                                                             │
└─────────────────────────────────────────────────────────────┘

E2E Test Script Template

#!/bin/bash
# E2E Test Script Template

API_URL="http://localhost:8000"
TOKEN="your-test-token"

PASS_COUNT=0
FAIL_COUNT=0
SKIP_COUNT=0

GREEN='\033[0;32m'
RED='\033[0;31m'
YELLOW='\033[0;33m'
NC='\033[0m'

test_feature_action() {
    echo -n "Testing: Feature action... "

    response=$(curl -s -X POST "$API_URL/api/v1/feature/action" \
        -H "Authorization: Bearer $TOKEN" \
        -H "Content-Type: application/json" \
        -d '{"param": "value"}')

    if [[ "$response" == *"expected_result"* ]]; then
        echo -e "${GREEN}✅ PASS${NC}"
        ((PASS_COUNT++))
    else
        echo -e "${RED}❌ FAIL${NC}"
        echo "Response: $response"
        ((FAIL_COUNT++))
    fi
}

# Run all tests
test_feature_action
# ... more tests

# Summary
echo ""
echo "═══════════════════════════════════════"
echo "Test Results:"
echo -e "  ${GREEN}✅ PASS: $PASS_COUNT${NC}"
echo -e "  ${RED}❌ FAIL: $FAIL_COUNT${NC}"
echo -e "  ${YELLOW}⏭️  SKIP: $SKIP_COUNT${NC}"
echo "═══════════════════════════════════════"

Test Cycle Documentation Template

# Feature Test Results - Cycle N

## Summary
- **Date**: YYYY-MM-DD
- **Feature**: {feature name}
- **Pass Rate**: N%
- **Tests**: X passed / Y total

## Results

| Test Case | Status | Notes |
|-----------|--------|-------|
| Test 1 | ✅ | |
| Test 2 | ❌ | {error description} |
| Test 3 | ⏭️ | {skip reason} |

## Bugs Found

### BUG-001: {Title}
- **Root Cause**: {description}
- **Fix**: {what was changed}
- **Files**: `path/to/file.py:123`

## Next Cycle Plan
- {what to test next}

Checklist

Logging Infrastructure

• JSON log format applied
• Request ID generation and propagation
• Log level settings per environment
• Docker logging configuration

Backend Logging

• Logging middleware implemented
• All API calls logged (including 200 OK)
• Business logic logging
• Detailed error logging

Frontend Logging

• Logger module implemented
• API client integration
• Error boundary logging

QA Workflow

• Docker Compose configured
• Real-time monitoring ready
• Issue documentation template ready

Auto-Apply Rules

When Building Logging Infrastructure

When implementing API/Backend:

1. Suggest logging middleware creation
2. Suggest JSON format logger setup
3. Add Request ID generation/propagation logic

When implementing Frontend:

1. Suggest Logger module creation
2. Suggest logging integration with API client
3. Suggest including Request ID header

When Performing QA

On test request:

1. Guide to run docker compose logs -f
2. Request manual UX testing from user
3. Real-time log monitoring
4. Document issues immediately when detected
5. Provide fix suggestions

Issue Detection Thresholds

Required Logging Locations

Backend (FastAPI/Express)

✅ Request start (method, path, params)
✅ Request complete (status, duration_ms)
✅ Major business logic steps
✅ Detailed info on errors
✅ Before/after external API calls
✅ DB queries (in development)

Frontend (Next.js/React)

✅ API call start
✅ API response received (status, duration)
✅ Detailed info on errors
✅ Important user actions