Claude-code-plugins-plus-skills serpapi-reference-architecture

install
source · Clone the upstream repo
git clone https://github.com/jeremylongshore/claude-code-plugins-plus-skills
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/jeremylongshore/claude-code-plugins-plus-skills "$T" && mkdir -p ~/.claude/skills && cp -r "$T/plugins/saas-packs/serpapi-pack/skills/serpapi-reference-architecture" ~/.claude/skills/jeremylongshore-claude-code-plugins-plus-skills-serpapi-reference-architecture && rm -rf "$T"
manifest: plugins/saas-packs/serpapi-pack/skills/serpapi-reference-architecture/SKILL.md
source content

SerpApi Reference Architecture

Overview

Production architecture for search-powered applications using SerpApi. Core components: cached search service, multi-engine abstraction, SERP monitoring pipeline, and credit budget management.

Architecture Diagram

┌──────────────────────────────────┐
│          API Layer               │
│  /search  /track  /health        │
├──────────────────────────────────┤
│        Search Service            │
│  Multi-engine  Caching  Parsing  │
├──────────────────────────────────┤
│       SerpApi Client             │
│  Rate Limiting  Retry  Archive   │
├──────────────────────────────────┤
│       Infrastructure             │
│  Redis Cache  PostgreSQL  Cron   │
└──────────────────────────────────┘
         │
         ▼
┌──────────────────────────────────┐
│        SerpApi REST API          │
│  google  youtube  bing  news     │
│  1 credit/search, 100-50K/mo     │
└──────────────────────────────────┘

Project Structure

search-service/
├── src/
│   ├── serpapi/
│   │   ├── client.ts          # Cached search with rate limiting
│   │   ├── engines.ts         # Engine-specific param mapping
│   │   └── types.ts           # Typed result interfaces
│   ├── services/
│   │   ├── search.ts          # Multi-engine search facade
│   │   ├── tracking.ts        # Keyword rank tracking
│   │   └── credits.ts         # Usage monitoring
│   ├── api/
│   │   ├── search.ts          # /search proxy endpoint
│   │   └── health.ts          # /health with credit check
│   └── jobs/
│       └── rank-tracker.ts    # Daily keyword monitoring
├── tests/
│   ├── fixtures/              # Recorded SerpApi responses
│   └── search.test.ts         # Fixture-based tests
└── config/

Key Components

Search Service Facade

class SearchService {
  constructor(private client: CachedSerpApiClient, private db: Database) {}

  async search(query: string, options?: { engine?: string; num?: number }) {
    const engine = options?.engine || 'google';
    const result = await this.client.cachedSearch({
      engine, q: query, num: options?.num || 5,
    });

    // Normalize across engines
    return {
      results: result.organic_results || result.video_results || [],
      answer_box: result.answer_box || null,
      knowledge_graph: result.knowledge_graph || null,
      search_id: result.search_metadata.id,
      cached: result._cached || false,
    };
  }

  async trackKeyword(keyword: string, domain: string) {
    const result = await this.client.cachedSearch({
      engine: 'google', q: keyword, num: 100,
    });
    const position = result.organic_results?.findIndex(
      (r: any) => r.link?.includes(domain)
    );
    await this.db.saveRanking(keyword, domain, position >= 0 ? position + 1 : null);
  }
}

Credit Budget Manager

class CreditBudget {
  async check(): Promise<{ ok: boolean; remaining: number }> {
    const account = await fetch(
      `https://serpapi.com/account.json?api_key=${process.env.SERPAPI_API_KEY}`
    ).then(r => r.json());

    return {
      ok: account.plan_searches_left > 100,
      remaining: account.plan_searches_left,
    };
  }
}

Error Handling

ComponentFailureRecovery
Search APICredits exhaustedReturn cached results, alert ops
CacheRedis downFall through to API (graceful degradation)
Rank trackerQuery failsSkip and retry next cycle
Health checkAPI unreachableReport degraded status

Resources

Next Steps

See individual skill docs for deep-dives on each component.