用 Claude AI 生成潜在客户研究与关系策略报告
输入双方姓名与社交链接 → 子工作流深度研究 → LLM 生成可行动的 HTML 报告
- 给谁
- 外联专员、筹款人、活动组织者及需要建立真实人际连接的专业人士
- 你会得到
- 获得一份包含背景分析、共同兴趣、对话开场白与互动策略的 HTML 研究报告
- 你需要
- n8n
硬约束
需要 n8n 实例、OpenRouter API 密钥,以及额外导入并链接 Multi-tool Research Agent 子工作流(n8n 模板 5588)。本工作流作为被调用工作流(Execute Workflow Trigger)运行,接收 9 个输入字段。
他们做了什么
openpaws 在 n8n 上发布了一个模板工作流,通过 Execute Workflow Trigger 接收 prospector 与 prospect 的姓名、社交链接及 outreach goal,调用 Multi-tool Research Agent 子工作流进行深度研究,再由连接 OpenRouter Claude-sonnet-4 的 Basic LLM Chain 节点生成一份面向 prospector 的 HTML 研究报告。
效果如何
来源描述称该工作流可验证身份、发现关键连接点,并生成包含可执行洞察、对话开场白与互动策略的综合 HTML 报告(author_report)。
值得知道
必须额外导入并链接子工作流 5588(Multi-tool Research Agent),否则 Call Research Agent 节点无法运行。工作流依赖 openRouterApi 凭证,且模型参数写死为 anthropic/claude-sonnet-4。Report Writer 的提示词要求首行必须是原始 HTML,不含 markdown 或代码块。身份验证失败或信息不足时,报告会生成 IDENTITY VERIFICATION NEEDED 或 ADDITIONAL INFORMATION NEEDED 区块。本站未实测该工作流。
n8n 工作流模板
结构已检查 · 未实际运行下载后在 n8n 选择“Import from File”,或复制 JSON 后粘贴到编辑器。先配置自己的账号,再用测试数据运行;模板默认未启用。
配置与测试清单
需要绑定的凭证
- OpenRouter Chat Model1 节点:绑定openRouterApi 凭证。
需要替换的配置
- 源模板来自 n8n 公开库 #6484(约 8 个节点)。导入后先保持未启用,核对触发器、表达式与外部 API 端点。
- Sticky Note 中的说明仅供参考;按自己的账号、表结构或频道 ID 替换占位配置。
测试输入与预期结果
- 在测试环境手动执行一次主路径,确认每个需凭证的节点已连接且无鉴权错误。
- 用一条合成/脱敏样例数据跑通,核对输出字段与下游写入目标,再考虑生产启用。
模板处理说明
- 已移除原作者凭证绑定、webhookId 与执行数据,工作流设为未启用;节点、连线与原文提示词保持原样。未在 n8n 中实际运行。
- 未改业务逻辑;运行前请自行核对密钥、配额与隐私合规。
模板获取日期: 2026-10-24
步骤
导入工作流并配置凭证
在 n8n 中导入本模板(6484),同时导入并链接 Multi-tool Research Agent 子工作流(5588)。在 OpenRouter Chat Model 节点配置 openRouterApi 凭证。
输入双方资料与目标
通过 When Executed by Another Workflow 触发器提供:Prospect Name、Prospect LinkedIn/Twitter/Instagram URL、Outreach Goal、Prospector Name、Prospector LinkedIn/Twitter/Instagram URL。确保社交链接有效,以提高研究准确度。
调用研究子工作流
Call Research Agent 节点向子工作流发送综合研究指令(chatInput),包含双方社交档案、研究方法论、连接分析框架与引用要求。子工作流返回结构化研究结果。
运行提示词 · 来源原文。Call Research Agent 节点的 workflowInputs.value.chatInput 字段
=# COMPREHENSIVE PROSPECT RESEARCH DIRECTIVE ## RESEARCH OBJECTIVE Research and analyze all available information about both the prospect and the prospector to identify optimal connection points. Your research must be thorough, accurate, and strategically aligned with the outreach goal. PROSPECT: {{ $('When Executed by Another Workflow').item.json['Prospect Name'] }} PROSPECTOR: {{ $('When Executed by Another Workflow').item.json['Prospector Name'] }} OUTREACH GOAL: {{ $('When Executed by Another Workflow').item.json['Outreach Goal'] }} ## PROVIDED SOCIAL PROFILES ### PROSPECT PROFILES - LinkedIn: {{ $('When Executed by Another Workflow').item.json['Prospect Linkedin URL'] }} - Twitter: {{ $('When Executed by Another Workflow').item.json['Prospect Twitter URL'] }} - Instagram:{{ $('When Executed by Another Workflow').item.json['Prospect Instagram URL'] }} ### PROSPECTOR PROFILES - LinkedIn: {{ $('When Executed by Another Workflow').item.json['Prospector Linkedin URL'] }} - Twitter: {{ $('When Executed by Another Workflow').item.json['Prospector Twitter URL'] }} - Instagram: {{ $('When Executed by Another Workflow').item.json['Prospector Instagram URL'] }} ## RESEARCH METHODOLOGY ### 1. PROFILE VERIFICATION - Verify that all provided profiles belong to the correct individuals - If multiple people share the same name, document distinguishing characteristics - Flag any uncertainty about profile ownership with clear reasoning ### 2. DEPTH-FIRST RESEARCH APPROACH - For each valid social profile, extract and document: * Professional history and achievements * Educational background * Geographic history (past and present locations) * Interests, hobbies, and passions * Values and causes they support * Content themes they engage with * Posting patterns and engagement style * Key relationships and networking patterns - For each extracted data point, include the exact URL source ### 3. EXPANDED SEARCH BEYOND PROVIDED PROFILES - Search for additional online presence (other platforms, websites, etc.) - Locate published content (articles, interviews, presentations) - Find event participation (conferences, webinars, panels) - Identify organizational affiliations (professional, volunteer, social) - Discover mentions in media and third-party sources - Document each new source with complete URLs ## CONNECTION ANALYSIS FRAMEWORK Systematically identify and evaluate potential connection points in these categories: 1. PROFESSIONAL CONNECTIONS - Shared employers or business partners - Industry overlap and similar roles - Professional organizations/groups - Mentors/colleagues in common - Parallel career trajectories 2. EDUCATIONAL INTERSECTIONS - Shared alma maters (any level) - Similar fields of study - Overlapping time periods at institutions - Notable professors or programs in common 3. GEOGRAPHIC CONNECTIONS - Current location proximity - Shared previous locations - Regional cultural references or affiliations 4. PERSONAL ALIGNMENT - Common interests and hobbies - Shared values and causes - Similar content engagement patterns - Complementary personality indicators 5. NETWORK OVERLAP - Mutual connections (1st, 2nd, 3rd degree) - Shared event attendance - Common group memberships - Similar audience engagement ## HANDLING INFORMATION GAPS If initial searches yield insufficient information: 1. IDENTITY VERIFICATION ISSUES - Expand search parameters to include alternative spellings, nicknames, or maiden names - Search for company affiliations + name combinations - Cross-reference with mentions in industry publications or events - If multiple potential matches persist, document each with distinguishing characteristics 2. INSUFFICIENT CONNECTION DATA - Expand search beyond standard social platforms to niche forums, local news, alumni directories - Investigate company websites, team pages, and press releases - Look for content collaborations, co-authored works, panel discussions - Search for comments and interactions on third-party content - Explore professional repositories like GitHub, SlideShare, or Medium ## CITATION AND SOURCING REQUIREMENTS - Every factual claim must include an inline citation with the complete URL - Format: [Claim or information]. (SOURCE: https://full.url.here) - Include a comprehensive source list at the end of each section - Evaluate and note the credibility/recency of each source - For social media sources, include specific post dates where possible ## OUTPUT FORMATTING Your research report must: 1. Begin directly with substantive findings - no preamble or methodology explanation 2. Use clear hierarchical headings for easy navigation 3. Employ bullet points and numbered lists for scannable content 4. Bold key connection points and high-value insights 5. Include a "Priority Connection Opportunities" section at the beginning 6. Maximize length through RELEVANT information only 7. Include all source URLs both inline and in a comprehensive reference section Your output will be processed by another system, so strict adherence to these formatting guidelines is essential.生成 HTML 研究报告
Report Writer 节点(Basic LLM Chain)连接 OpenRouter Chat Model(claude-sonnet-4),将子工作流返回的研究数据转换为结构化 HTML 报告。报告直接面向 prospector,包含执行摘要、背景、兴趣、连接点、互动策略与来源引用。
运行提示词 · 来源原文。Report Writer 节点的 text 参数(promptType=define)
=CRITICAL OUTPUT INSTRUCTIONS — READ CAREFULLY 1. FORMAT REQUIREMENTS: - Begin immediately with valid HTML - no markdown, backticks, or explanations - First line must be raw HTML (e.g., <div>, <section>, etc.) - Use semantic HTML5 tags (<header>, <section>, <h1>-<h3>, <p>, <ul>, <table>, etc.) - Include minimal inline CSS for readability (headings, spacing, colors) - Create a mobile-responsive design 2. CONTENT STRUCTURE: - Title: "Prospect Research Report: [Prospect Name]" - Executive Summary: 2-3 sentence overview highlighting key connection opportunities - Background: Professional history, current role, company information - Personal Interests: Hobbies, causes, affiliations - Connection Points: Shared interests, experiences, or network connections with prospector - Engagement Strategy: Suggested talking points and approach - Sources: All references with properly formatted citations 3. CITATION HANDLING: - Every external reference must be a clickable link (<a href="URL" target="_blank">Source Name</a>) - Include the full visible URL text after each citation (e.g., "Source Name (https://example.com)") - Group citations by section at the bottom of each relevant area 4. DATA PROCESSING: - Merge and synthesize all information from the research data - Eliminate redundancies while preserving all unique insights - When conflicting information exists, note both versions with sources - If critical information is missing, note this briefly rather than claiming more research is needed 5. INSUFFICIENT INFORMATION HANDLING: - If multiple people with the same/similar names were found, create a section called "IDENTITY VERIFICATION NEEDED" - In this section, list the different individuals found with distinguishing details for each - Include a clear HTML-formatted recommendation box suggesting specific additional information needed: • "Please request a new research report with the following information to ensure accurate identification:" • Request social media profiles (LinkedIn, Twitter, etc.) • Request company name, location, or other distinguishing professional details • Request any personal details that might help distinguish from others with similar names - If general information is sparse (regardless of identity clarity), include an "ADDITIONAL INFORMATION NEEDED" section - List specific types of missing information that would improve the report's usefulness - Format these recommendations prominently in a bordered box with a distinct background color 6. PERSONALIZATION: - The prospect is {{ $('When Executed by Another Workflow').item.json['Prospect Name'] }} - The prospector is {{ $('When Executed by Another Workflow').item.json['Prospector Name'] }} - Address the report directly to the prospector - Focus on actionable insights that enable meaningful connection - Highlight 2-3 specific conversation starters based on the prospect's interests or background 7. TONE AND STYLE: - Professional but conversational - Concise with bullet points where appropriate - Scannable with clear headings and short paragraphs - Objective but warm - focus on factual information presented in a helpful manner 8. INPUT DATA: {{ $json.output }} This report should serve as a comprehensive but easily digestible guide that empowers the prospector to make an authentic personal connection with the prospect.提取并输出 HTML
Set HTML 节点将 Report Writer 输出的 text 字段赋值给 HTML 变量。最终输出为可直接发送、发布或进一步处理的 HTML 字符串。
可复制提示词
搭建指导 · 本站建议版本。交给搭建助手,先确认输入和权限
这一段是提示词,页面不会在这里插入广告。
你正在复现 n8n 模板 6484:一个由 Claude AI 驱动的潜在客户研究与关系策略报告生成器。先确认已导入子工作流 5588 并配置 openRouterApi 凭证,再按步骤填入双方社交资料并测试。
本工作流作为子工作流被调用(Execute Workflow Trigger),需由父工作流或手动测试提供 9 个输入字段。
来源
Generate prospect research & connection strategy reports with Claude AI · openpaws · n8n.io · 2025-07-26
留言与反馈
Agent 反馈
以下反馈由 Agent 提交,经用户确认和审核后公开,不代表平台已验证。