{"skill":{"slug":"openclaw-aisa-search-website-academic-tavily-serp-exa","displayName":"Multi-source retrieval with confidence scoring - web, academic, and Tavily in one unified API","summary":"Intelligent search for agents. Multi-source retrieval with confidence scoring - web, academic, and Tavily in one unified API.","description":"---\r\nname: openclaw-search\r\ndescription: \"Intelligent search for agents. Multi-source retrieval with confidence scoring - web, academic, and Tavily in one unified API.\"\r\nhomepage: https://openclaw.ai\r\nmetadata: {\"openclaw\":{\"emoji\":\"🔍\",\"requires\":{\"bins\":[\"curl\",\"python3\"],\"env\":[\"AISA_API_KEY\"]},\"primaryEnv\":\"AISA_API_KEY\"}}\r\n---\r\n\r\n# OpenClaw Search 🔍\r\n\r\n**Intelligent search for autonomous agents. Powered by AIsa.**\r\n\r\nOne API key. Multi-source retrieval. Confidence-scored answers.\r\n\r\n> Inspired by [AIsa Verity](https://github.com/AIsa-team/verity) - A next-generation search agent with trust-scored answers.\r\n\r\n## 🔥 What Can You Do?\r\n\r\n### Research Assistant\r\n```\r\n\"Search for the latest papers on transformer architectures from 2024-2025\"\r\n```\r\n\r\n### Market Research\r\n```\r\n\"Find all web articles about AI startup funding in Q4 2025\"\r\n```\r\n\r\n### Competitive Analysis\r\n```\r\n\"Search for reviews and comparisons of RAG frameworks\"\r\n```\r\n\r\n### News Aggregation\r\n```\r\n\"Get the latest news about quantum computing breakthroughs\"\r\n```\r\n\r\n### Deep Dive Research\r\n```\r\n\"Smart search combining web and academic sources on 'autonomous agents'\"\r\n```\r\n\r\n## Quick Start\r\n\r\n```bash\r\nexport AISA_API_KEY=\"your-key\"\r\n```\r\n\r\n---\r\n\r\n## 🏗️ Architecture: Multi-Stage Orchestration\r\n\r\nOpenClaw Search employs a **Two-Phase Retrieval Strategy** for comprehensive results:\r\n\r\n### Phase 1: Discovery (Parallel Retrieval)\r\n\r\nQuery 4 distinct search streams simultaneously:\r\n- **Scholar**: Deep academic retrieval\r\n- **Web**: Structured web search\r\n- **Smart**: Intelligent mixed-mode search\r\n- **Tavily**: External validation signal\r\n\r\n### Phase 2: Reasoning (Meta-Analysis)\r\n\r\nUse **AIsa Explain** to perform meta-analysis on search results, generating:\r\n- Confidence scores (0-100)\r\n- Source agreement analysis\r\n- Synthesized answers\r\n\r\n```\r\n┌─────────────────────────────────────────────────────────────┐\r\n│                      User Query                              │\r\n└─────────────────────────────────────────────────────────────┘\r\n                              │\r\n              ┌───────────────┼───────────────┐\r\n              ▼               ▼               ▼\r\n        ┌─────────┐     ┌─────────┐     ┌─────────┐\r\n        │ Scholar │     │   Web   │     │  Smart  │\r\n        └─────────┘     └─────────┘     └─────────┘\r\n              │               │               │\r\n              └───────────────┼───────────────┘\r\n                              ▼\r\n                    ┌─────────────────┐\r\n                    │  AIsa Explain   │\r\n                    │ (Meta-Analysis) │\r\n                    └─────────────────┘\r\n                              │\r\n                              ▼\r\n                    ┌─────────────────┐\r\n                    │ Confidence Score│\r\n                    │  + Synthesis    │\r\n                    └─────────────────┘\r\n```\r\n\r\n---\r\n\r\n## Core Capabilities\r\n\r\n### Web Search\r\n\r\n```bash\r\n# Basic web search\r\ncurl -X POST \"https://api.aisa.one/apis/v1/scholar/search/web?query=AI+frameworks&max_num_results=10\" \\\r\n  -H \"Authorization: Bearer $AISA_API_KEY\"\r\n\r\n# Full text search (with page content)\r\ncurl -X POST \"https://api.aisa.one/apis/v1/search/full?query=latest+AI+news&max_num_results=10\" \\\r\n  -H \"Authorization: Bearer $AISA_API_KEY\"\r\n```\r\n\r\n### Academic/Scholar Search\r\n\r\n```bash\r\n# Search academic papers\r\ncurl -X POST \"https://api.aisa.one/apis/v1/scholar/search/scholar?query=transformer+models&max_num_results=10\" \\\r\n  -H \"Authorization: Bearer $AISA_API_KEY\"\r\n\r\n# With year filter\r\ncurl -X POST \"https://api.aisa.one/apis/v1/scholar/search/scholar?query=LLM&max_num_results=10&as_ylo=2024&as_yhi=2025\" \\\r\n  -H \"Authorization: Bearer $AISA_API_KEY\"\r\n```\r\n\r\n### Smart Search (Web + Academic Combined)\r\n\r\n```bash\r\n# Intelligent hybrid search\r\ncurl -X POST \"https://api.aisa.one/apis/v1/scholar/search/smart?query=machine+learning+optimization&max_num_results=10\" \\\r\n  -H \"Authorization: Bearer $AISA_API_KEY\"\r\n```\r\n\r\n### Tavily Integration (Advanced)\r\n\r\n```bash\r\n# Tavily search\r\ncurl -X POST \"https://api.aisa.one/apis/v1/tavily/search\" \\\r\n  -H \"Authorization: Bearer $AISA_API_KEY\" \\\r\n  -H \"Content-Type: application/json\" \\\r\n  -d '{\"query\":\"latest AI developments\"}'\r\n\r\n# Extract content from URLs\r\ncurl -X POST \"https://api.aisa.one/apis/v1/tavily/extract\" \\\r\n  -H \"Authorization: Bearer $AISA_API_KEY\" \\\r\n  -H \"Content-Type: application/json\" \\\r\n  -d '{\"urls\":[\"https://example.com/article\"]}'\r\n\r\n# Crawl web pages\r\ncurl -X POST \"https://api.aisa.one/apis/v1/tavily/crawl\" \\\r\n  -H \"Authorization: Bearer $AISA_API_KEY\" \\\r\n  -H \"Content-Type: application/json\" \\\r\n  -d '{\"url\":\"https://example.com\",\"max_depth\":2}'\r\n\r\n# Site map\r\ncurl -X POST \"https://api.aisa.one/apis/v1/tavily/map\" \\\r\n  -H \"Authorization: Bearer $AISA_API_KEY\" \\\r\n  -H \"Content-Type: application/json\" \\\r\n  -d '{\"url\":\"https://example.com\"}'\r\n```\r\n\r\n### Explain Search Results (Meta-Analysis)\r\n\r\n```bash\r\n# Generate explanations with confidence scoring\r\ncurl -X POST \"https://api.aisa.one/apis/v1/scholar/explain\" \\\r\n  -H \"Authorization: Bearer $AISA_API_KEY\" \\\r\n  -H \"Content-Type: application/json\" \\\r\n  -d '{\"results\":[...],\"language\":\"en\",\"format\":\"summary\"}'\r\n```\r\n\r\n---\r\n\r\n## 📊 Confidence Scoring Engine\r\n\r\nUnlike standard RAG systems, OpenClaw Search evaluates credibility and consensus:\r\n\r\n### Scoring Rubric\r\n\r\n| Factor | Weight | Description |\r\n|--------|--------|-------------|\r\n| **Source Quality** | 40% | Academic > Smart/Web > External |\r\n| **Agreement Analysis** | 35% | Cross-source consensus checking |\r\n| **Recency** | 15% | Newer sources weighted higher |\r\n| **Relevance** | 10% | Query-result semantic match |\r\n\r\n### Score Interpretation\r\n\r\n| Score | Confidence Level | Meaning |\r\n|-------|-----------------|---------|\r\n| 90-100 | Very High | Strong consensus across academic and web sources |\r\n| 70-89 | High | Good agreement, reliable sources |\r\n| 50-69 | Medium | Mixed signals, verify independently |\r\n| 30-49 | Low | Conflicting sources, use caution |\r\n| 0-29 | Very Low | Insufficient or contradictory data |\r\n\r\n---\r\n\r\n## Python Client\r\n\r\n```bash\r\n# Web search\r\npython3 {baseDir}/scripts/search_client.py web --query \"latest AI news\" --count 10\r\n\r\n# Academic search\r\npython3 {baseDir}/scripts/search_client.py scholar --query \"transformer architecture\" --count 10\r\npython3 {baseDir}/scripts/search_client.py scholar --query \"LLM\" --year-from 2024 --year-to 2025\r\n\r\n# Smart search (web + academic)\r\npython3 {baseDir}/scripts/search_client.py smart --query \"autonomous agents\" --count 10\r\n\r\n# Full text search\r\npython3 {baseDir}/scripts/search_client.py full --query \"AI startup funding\"\r\n\r\n# Tavily operations\r\npython3 {baseDir}/scripts/search_client.py tavily-search --query \"AI developments\"\r\npython3 {baseDir}/scripts/search_client.py tavily-extract --urls \"https://example.com/article\"\r\n\r\n# Multi-source search with confidence scoring\r\npython3 {baseDir}/scripts/search_client.py verity --query \"Is quantum computing ready for enterprise?\"\r\n```\r\n\r\n---\r\n\r\n## API Endpoints Reference\r\n\r\n| Endpoint | Method | Description |\r\n|----------|--------|-------------|\r\n| `/scholar/search/web` | POST | Web search with structured results |\r\n| `/scholar/search/scholar` | POST | Academic paper search |\r\n| `/scholar/search/smart` | POST | Intelligent hybrid search |\r\n| `/scholar/explain` | POST | Generate result explanations |\r\n| `/search/full` | POST | Full text search with content |\r\n| `/search/smart` | POST | Smart web search |\r\n| `/tavily/search` | POST | Tavily search integration |\r\n| `/tavily/extract` | POST | Extract content from URLs |\r\n| `/tavily/crawl` | POST | Crawl web pages |\r\n| `/tavily/map` | POST | Generate site maps |\r\n\r\n---\r\n\r\n## Search Parameters\r\n\r\n| Parameter | Type | Description |\r\n|-----------|------|-------------|\r\n| query | string | Search query (required) |\r\n| max_num_results | integer | Max results (1-100, default 10) |\r\n| as_ylo | integer | Year lower bound (scholar only) |\r\n| as_yhi | integer | Year upper bound (scholar only) |\r\n\r\n---\r\n\r\n## 🚀 Building a Verity-Style Agent\r\n\r\nWant to build your own confidence-scored search agent? Here's the pattern:\r\n\r\n### 1. Parallel Discovery\r\n\r\n```python\r\nimport asyncio\r\n\r\nasync def discover(query):\r\n    \"\"\"Phase 1: Parallel retrieval from multiple sources.\"\"\"\r\n    tasks = [\r\n        search_scholar(query),\r\n        search_web(query),\r\n        search_smart(query),\r\n        search_tavily(query)\r\n    ]\r\n    results = await asyncio.gather(*tasks)\r\n    return {\r\n        \"scholar\": results[0],\r\n        \"web\": results[1],\r\n        \"smart\": results[2],\r\n        \"tavily\": results[3]\r\n    }\r\n```\r\n\r\n### 2. Confidence Scoring\r\n\r\n```python\r\ndef score_confidence(results):\r\n    \"\"\"Calculate deterministic confidence score.\"\"\"\r\n    score = 0\r\n    \r\n    # Source quality (40%)\r\n    if results[\"scholar\"]:\r\n        score += 40 * len(results[\"scholar\"]) / 10\r\n    \r\n    # Agreement analysis (35%)\r\n    claims = extract_claims(results)\r\n    agreement = analyze_agreement(claims)\r\n    score += 35 * agreement\r\n    \r\n    # Recency (15%)\r\n    recency = calculate_recency(results)\r\n    score += 15 * recency\r\n    \r\n    # Relevance (10%)\r\n    relevance = calculate_relevance(results, query)\r\n    score += 10 * relevance\r\n    \r\n    return min(100, score)\r\n```\r\n\r\n### 3. Synthesis\r\n\r\n```python\r\nasync def synthesize(query, results, score):\r\n    \"\"\"Generate final answer with citations.\"\"\"\r\n    explanation = await explain_results(results)\r\n    return {\r\n        \"answer\": explanation[\"summary\"],\r\n        \"confidence\": score,\r\n        \"sources\": explanation[\"citations\"],\r\n        \"claims\": explanation[\"claims\"]\r\n    }\r\n```\r\n\r\nFor a complete implementation, see [AIsa Verity](https://github.com/AIsa-team/verity).\r\n\r\n---\r\n\r\n## Pricing\r\n\r\n| API | Cost |\r\n|-----|------|\r\n| Web search | ~$0.001 |\r\n| Scholar search | ~$0.002 |\r\n| Smart search | ~$0.002 |\r\n| Tavily search | ~$0.002 |\r\n| Explain | ~$0.003 |\r\n\r\nEvery response includes `usage.cost` and `usage.credits_remaining`.\r\n\r\n---\r\n\r\n## Get Started\r\n\r\n1. Sign up at [aisa.one](https://aisa.one)\r\n2. Get your API key\r\n3. Add credits (pay-as-you-go)\r\n4. Set environment variable: `export AISA_API_KEY=\"your-key\"`\r\n\r\n## Full API Reference\r\n\r\nSee [API Reference](https://aisa.mintlify.app/api-reference/introduction) for complete endpoint documentation.\r\n\r\n## Resources\r\n\r\n- [AIsa Verity](https://github.com/AIsa-team/verity) - Reference implementation of confidence-scored search agent\r\n- [AIsa Documentation](https://aisa.mintlify.app) - Complete API documentation\r\n","tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":1626,"installsAllTime":61,"installsCurrent":4,"stars":4,"versions":1},"createdAt":1770385462220,"updatedAt":1778486065752},"latestVersion":{"version":"1.0.0","createdAt":1770385462220,"changelog":"- Initial release of OpenClaw Search: a unified API for intelligent, multi-source retrieval (web, academic, Tavily).\n- Provides confidence-scored answers through meta-analysis of retrieved results.\n- Supports parallel search via web, academic, smart, and Tavily streams.\n- Includes Python and curl client examples for all major features and endpoints.\n- Features confidence scoring engine with transparency on source quality, recency, and consensus.\n- Full API reference with endpoints and required parameters included.","license":null},"metadata":{"setup":[{"key":"AISA_API_KEY","required":true}],"os":null,"systems":null},"owner":{"handle":"0xjordansg-yolo","userId":"s17eqxzhby6aqcv44tccm6dqh583fznj","displayName":"0xjordansg-yolo","image":"https://avatars.githubusercontent.com/u/249538244?v=4"},"moderation":null}