#!/usr/bin/env python3 """临时脚本：从今日搜索结果插入新场景""" import sqlite3 import json import uuid from datetime import datetime # 新发现的场景数据 new_scenes = [ { "title": "Autonomous Customer Service Resolution", "description": "AI agents handle end-to-end customer service including ticket resolution, refunds, escalations, and omnichannel support. Saves teams 40+ hours monthly.", "categories": "Customer Service,Support", "tags": "autonomous,ticket-resolution,refunds,omnichannel", "key_technologies": "Agentic AI,Multi-Agent Systems", "innovation_score": 85, "source_platform": "Web Search", "impact": "Saves teams 40+ hours monthly" }, { "title": "Invoice Matching and Reconciliation", "description": "AI automates invoice matching, reconciliation, and expense auditing in finance operations. Accelerates processes by 30–50%.", "categories": "Finance,Operations", "tags": "finance,automation,reconciliation,auditing", "key_technologies": "AI Agents,RPA Integration", "innovation_score": 80, "source_platform": "Web Search", "impact": "Accelerates processes by 30–50%" }, { "title": "Multi-Agent Security Compliance", "description": "Agents collaborate for anomaly/threat detection, policy enforcement, and incident remediation with proactive risk reduction.", "categories": "Security,Compliance,IT", "tags": "security,threat-detection,compliance,incident-response", "key_technologies": "Multi-Agent Systems,Security AI", "innovation_score": 90, "source_platform": "Web Search", "impact": "Enables proactive risk reduction" }, { "title": "Agentic Lead Generation and Qualification", "description": "AI agents scrape LinkedIn/Google Maps for prospects, enrich leads with contact info, and generate personalized outreach emails.", "categories": "Sales,Marketing", "tags": "lead-generation,prospecting,outreach,automation", "key_technologies": "Web Scraping Agents,Email Generation", "innovation_score": 88, "source_platform": "Web Search", "impact": "2–3x pipeline velocity improvement" }, { "title": "Predictive Maintenance and Supply Chain Optimization", "description": "AI analyzes sensor data and historical records to predict equipment failures, schedule maintenance, and optimize inventory/logistics.", "categories": "Manufacturing,Supply Chain", "tags": "predictive-maintenance,supply-chain,optimization", "key_technologies": "IoT Analytics,Predictive AI", "innovation_score": 87, "source_platform": "Web Search", "impact": "Minimizes downtime in manufacturing" }, { "title": "End-to-End Business Process Automation", "description": "Automates full workflows like customer onboarding, document verification, approvals, and closure by merging RPA with AI for decisions.", "categories": "Business Process,Operations", "tags": "onboarding,workflow-automation,RPA,AI", "key_technologies": "RPA,AI Decision Engines", "innovation_score": 85, "source_platform": "Web Search", "impact": "Reduces hand-offs and scales high-volume tasks" }, { "title": "Customer 360 Data Integration", "description": "Unifies fragmented data from CRM, marketing, and support platforms; detects duplicates and enriches profiles for real-time insights.", "categories": "Data Integration,Customer Experience", "tags": "data-unification,Customer-360,CRM,real-time", "key_technologies": "Data Integration,Real-time Analytics", "innovation_score": 82, "source_platform": "Web Search", "impact": "Enables personalized experiences and better sales targeting" }, { "title": "Competitive Intelligence and SEO Automation", "description": "Analyzes rivals' content activity, ads, or keywords; generates reports on patterns and evaluates AI visibility for strategic positioning.", "categories": "Marketing,Competitive Intelligence", "tags": "competitive-intelligence,SEO,analytics", "key_technologies": "Web Analytics,SEO Tools", "innovation_score": 80, "source_platform": "Web Search", "impact": "Strategic positioning and visibility insights" }, { "title": "Meeting Preparation and Knowledge Management", "description": "Agents research upcoming meetings across CRM, email, Slack, and web sources; or extract insights from articles/tickets into structured sheets.", "categories": "Productivity,Knowledge Management", "tags": "meeting-prep,research,knowledge-management", "key_technologies": "Multi-Source Research,Knowledge Graph", "innovation_score": 84, "source_platform": "Web Search", "impact": "Streamlines meeting preparation and knowledge extraction" }, { "title": "Automated Compliance, Audit, and Marketing Harmonization", "description": "Automates audit trails, report generation, data cleansing from multiple platforms, and real-time campaign optimization.", "categories": "Compliance,Audit,Marketing", "tags": "compliance,audit,marketing-optimization", "key_technologies": "Audit AI,Marketing Analytics", "innovation_score": 83, "source_platform": "Web Search", "impact": "Cuts review time by 70% and reduces regulatory risks" } ] # 数据库路径 db_path = "/root/.openclaw/workspace/projects/ai_usecase_collector/data/ai_usecases.db" # 连接数据库 conn = sqlite3.connect(db_path) cursor = conn.cursor() # 插入新场景 inserted_count = 0 skipped_count = 0 for scene in new_scenes: scene_id = str(uuid.uuid4()) current_time = datetime.now().isoformat() # 检查是否已存在相似标题 cursor.execute("SELECT id FROM scenes WHERE title = ?", (scene["title"],)) existing = cursor.fetchone() if existing: skipped_count += 1 print(f"跳过已存在: {scene['title']}") continue # 插入新场景 cursor.execute(""" INSERT INTO scenes ( id, title, description, source_platform, discovery_date, categories, tags, key_technologies, innovation_score, is_new, llm_summary, last_updated ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?) """, ( scene_id, scene["title"], scene["description"], scene["source_platform"], current_time, scene["categories"], scene["tags"], scene["key_technologies"], scene["innovation_score"], 1, # is_new json.dumps({"impact": scene.get("impact", "")}), current_time )) inserted_count += 1 print(f"插入新场景: {scene['title']}") # 记录日志 cursor.execute(""" INSERT INTO collection_log (run_date, source, items_found, items_new, status) VALUES (?, ?, ?, ?, ?) """, (current_time, "Daily Web Search", len(new_scenes), inserted_count, "completed")) conn.commit() conn.close() print(f"\n✅ 完成: 插入 {inserted_count} 个新场景，跳过 {skipped_count} 个已存在的场景")