使用Jina和Gemini的AI研究助手
高级
这是一个Content Creation, AI RAG领域的自动化工作流，包含 25 个节点。主要使用 Code, Wait, JinaAi, SplitInBatches, Agent 等节点。 使用Jina AI和Gemini 2.5 Flash创建全面的研究报告
前置要求
- •Google Gemini API Key
使用的节点 (25)
工作流预览
可视化展示节点连接关系，支持缩放和平移
导出工作流
复制以下 JSON 配置到 n8n 导入，即可使用此工作流
{
"id": "v9dRFBMYVwnQUhii",
"meta": {
"instanceId": "d06afcec45b6881aa3c2bd3e34579ea44e46952db235308e9371e2bb743e4b43",
"templateCredsSetupCompleted": true
},
"name": "使用 Jina 和 Gemini 的 AI 研究助手",
"tags": [],
"nodes": [
{
"id": "43de71f4-db4d-41aa-8294-abf2e0e984c8",
"name": "搜索网页",
"type": "n8n-nodes-base.jinaAi",
"position": [
20,
40
],
"parameters": {
"options": {
"outputFormat": ""
},
"operation": "search",
"searchQuery": "={{ $json.chatInput }}",
"requestOptions": {}
},
"credentials": {
"jinaAiApi": {
"id": "5ZbnfpUHVYg1i658",
"name": "Jina AI account"
}
},
"typeVersion": 1
},
{
"id": "8126e0ee-8472-4f1f-8921-1e2aded6682e",
"name": "当收到聊天消息时",
"type": "@n8n/n8n-nodes-langchain.chatTrigger",
"position": [
-300,
40
],
"webhookId": "19764858-2825-4532-98ce-8fc538cf9ed0",
"parameters": {
"options": {}
},
"typeVersion": 1.1
},
{
"id": "0c414f7e-d927-4a6b-a45c-412676d99986",
"name": "遍历项目",
"type": "n8n-nodes-base.splitInBatches",
"position": [
700,
40
],
"parameters": {
"options": {
"reset": false
}
},
"typeVersion": 3
},
{
"id": "0a8274f0-7ef4-4305-a3e7-81b34cd07ddc",
"name": "代码",
"type": "n8n-nodes-base.code",
"position": [
340,
40
],
"parameters": {
"language": "python",
"pythonCode": "extracted_urls = []\n\nfor item in _input.all():\n original_json = item.json\n\n if 'url' in original_json:\n extracted_urls.append({'url': original_json['url']})\n \nreturn extracted_urls"
},
"typeVersion": 2
},
{
"id": "540cae13-9d70-47b5-9be4-902abc7d6391",
"name": "读取URL内容",
"type": "n8n-nodes-base.jinaAi",
"position": [
1040,
200
],
"parameters": {
"url": "={{ $json.url }}",
"options": {},
"requestOptions": {}
},
"credentials": {
"jinaAiApi": {
"id": "5ZbnfpUHVYg1i658",
"name": "Jina AI account"
}
},
"typeVersion": 1
},
{
"id": "e282b54c-4230-449b-ab7c-96f0566e81f2",
"name": "等待",
"type": "n8n-nodes-base.wait",
"position": [
1980,
180
],
"webhookId": "d584a477-75aa-4d75-aa1b-0ef3c630f8cc",
"parameters": {
"amount": 1
},
"typeVersion": 1.1
},
{
"id": "37c6d91d-526e-4c9a-a1f3-4bbba63740bd",
"name": "结构化输出",
"type": "@n8n/n8n-nodes-langchain.outputParserStructured",
"position": [
1640,
340
],
"parameters": {
"jsonSchemaExample": "{\n\t\"source_url\": \"...\",\n\t\"summarized_content\": \"...\"\n}"
},
"typeVersion": 1.3
},
{
"id": "ef58119f-ab0e-4c99-a83d-d2a0dbc0ba17",
"name": "转换",
"type": "n8n-nodes-base.code",
"position": [
1060,
-180
],
"parameters": {
"language": "python",
"pythonCode": "all_outputs = []\nfor item in _input.all():\n current_item_data = item.json\n\n # Safely get the 'url' from the current item's data\n output = current_item_data.get('output')\n\n # Append the URL to the list if it exists\n if output:\n all_outputs.append(output)\n\n# Create the final single dictionary with all URLs under the \"urls\" key\nfinal_output_dict = {\n \"output\": all_outputs\n}\n\n# Return the final dictionary as a single n8n item\nreturn [{'json': final_output_dict}]"
},
"typeVersion": 2
},
{
"id": "154f148c-2f33-4d1d-afe2-2b75adfa16c5",
"name": "便签",
"type": "n8n-nodes-base.stickyNote",
"position": [
-420,
-160
],
"parameters": {
"width": 340,
"height": 380,
"content": "## 用户查询"
},
"typeVersion": 1
},
{
"id": "0bb609ef-222d-4cc6-8ffd-0145cea615f3",
"name": "便签1",
"type": "n8n-nodes-base.stickyNote",
"position": [
-60,
-160
],
"parameters": {
"color": 7,
"width": 260,
"height": 380,
"content": "## 搜索网页"
},
"typeVersion": 1
},
{
"id": "01ee087c-e26c-4b8f-9f3f-9598d0cd83d3",
"name": "便签2",
"type": "n8n-nodes-base.stickyNote",
"position": [
220,
-160
],
"parameters": {
"color": 2,
"width": 320,
"height": 380,
"content": "## 代码 (URL 转换)"
},
"typeVersion": 1
},
{
"id": "76f010e5-fc84-4066-a39a-7dd870ff9345",
"name": "便签3",
"type": "n8n-nodes-base.stickyNote",
"position": [
560,
-160
],
"parameters": {
"color": 6,
"width": 380,
"height": 380,
"content": "## 遍历所有 URL"
},
"typeVersion": 1
},
{
"id": "6cdc7a5f-b488-4b72-a0da-9351c7478ea1",
"name": "便签4",
"type": "n8n-nodes-base.stickyNote",
"position": [
960,
100
],
"parameters": {
"color": 3,
"width": 260,
"height": 480,
"content": "## 读取 URL 内容"
},
"typeVersion": 1
},
{
"id": "0d3c40cd-733b-4508-adc6-e34171e48ef1",
"name": "便签5",
"type": "n8n-nodes-base.stickyNote",
"position": [
1240,
100
],
"parameters": {
"color": 5,
"width": 600,
"height": 480,
"content": "## 汇总收集的内容"
},
"typeVersion": 1
},
{
"id": "30e03e88-16a5-4a77-bf19-74488042d88f",
"name": "### 需要帮助？",
"type": "n8n-nodes-base.stickyNote",
"position": [
1860,
100
],
"parameters": {
"color": 4,
"width": 340,
"height": 480,
"content": "## 避免速率限制"
},
"typeVersion": 1
},
{
"id": "52fd4d1c-63b6-44f0-b43f-30f092eadf7f",
"name": "## 试试看！",
"type": "n8n-nodes-base.stickyNote",
"position": [
960,
-480
],
"parameters": {
"width": 300,
"height": 480,
"content": "## 转换 (摘要聚合器)"
},
"typeVersion": 1
},
{
"id": "1572a6bd-ed9b-4461-9167-4b1d98bfc878",
"name": "GET 模型",
"type": "n8n-nodes-base.stickyNote",
"position": [
1280,
-480
],
"parameters": {
"color": 7,
"width": 480,
"height": 480,
"content": "## 生成研究报告"
},
"typeVersion": 1
},
{
"id": "f35efe92-7181-4283-8b93-858b807b1b40",
"name": "## 1. 创建新的自定义 OpenAI 凭据",
"type": "n8n-nodes-base.stickyNote",
"position": [
1780,
-480
],
"parameters": {
"width": 500,
"height": 480,
"content": "## 评估并生成最终报告"
},
"typeVersion": 1
},
{
"id": "d6d4b224-dd71-4bf7-be6f-b804002d48ac",
"name": "汇总器代理",
"type": "@n8n/n8n-nodes-langchain.agent",
"position": [
1420,
180
],
"parameters": {
"text": "=Summarize this information:\n\nSource url: {{ $json.url }}\nTitle of the page: {{ $json.title }}\nDescription: {{ $json.description }}\nContent: {{ $json.content }}\n\nInitial User query: {{ $('When chat message received').item.json.chatInput }}",
"options": {},
"promptType": "define",
"hasOutputParser": true
},
"typeVersion": 2
},
{
"id": "ef2cf6a3-3ea5-4874-898f-25ebc56997f1",
"name": "汇总器模型",
"type": "@n8n/n8n-nodes-langchain.lmChatGoogleGemini",
"position": [
1380,
340
],
"parameters": {
"options": {},
"modelName": "models/gemini-2.5-flash"
},
"credentials": {
"googlePalmApi": {
"id": "ngL4RI02lM8v6drj",
"name": "Google Gemini API"
}
},
"typeVersion": 1
},
{
"id": "4d26eb57-9e13-45e9-b365-c531189b0030",
"name": "生成器模型",
"type": "@n8n/n8n-nodes-langchain.lmChatGoogleGemini",
"position": [
1340,
-120
],
"parameters": {
"options": {},
"modelName": "models/gemini-2.5-flash"
},
"credentials": {
"googlePalmApi": {
"id": "ngL4RI02lM8v6drj",
"name": "Google Gemini API"
}
},
"typeVersion": 1
},
{
"id": "bc46ca58-24a6-4942-86e2-1c4c4df46c45",
"name": "生成器代理",
"type": "@n8n/n8n-nodes-langchain.agent",
"position": [
1380,
-320
],
"parameters": {
"text": "=## System Instructions\n\nYou are a specialized research AI agent designed to synthesize information from multiple sources and produce comprehensive, well-cited research documents. Your primary function is to analyze provided source URLs and their summarized contents to generate detailed research reports with proper inline citations.\n\n## Input \n\nInput with source_url and their summary : {\n\"1\": {\n \"source_url\": {{ $json.output[0].source_url }},\n \"summarized_content\": {{ $json.output[0].summarized_content }}\n },\n\"2\": {\n \"source_url\": {{ $json.output[1].source_url }},\n \"summarized_content\": {{ $json.output[1].summarized_content }}\n },\n\"3\": {\n \"source_url\": {{ $json.output[2] }},\n \"summarized_content\": {{ $json.output[2].summarized_content }}\n },\n\"4\": {\n \"source_url\": {{ $json.output[3].source_url }},\n \"summarized_content\": {{ $json.output[3].summarized_content }}\n },\n\"5\": {\n \"source_url\": {{ $json.output[4].source_url }},\n \"summarized_content\": {{ $json.output[4].summarized_content }}\n },\n\"6\": {\n \"source_url\": {{ $json.output[5].source_url }},\n \"summarized_content\": {{ $json.output[5].summarized_content }}\n },\n\"7\": {\n \"source_url\": {{ $json.output[6].source_url }},\n \"summarized_content\": {{ $json.output[6].summarized_content }}\n },\n\"8\": {\n \"source_url\": {{ $json.output[7].source_url }},\n \"summarized_content\": {{ $json.output[7].summarized_content }}\n },\n\"9\": {\n \"source_url\": {{ $json.output[8].source_url }},\n \"summarized_content\": {{ $json.output[8].summarized_content }}\n },\n\"10\": {\n \"source_url\": {{ $json.output[9].source_url }},\n \"summarized_content\": {{ $json.output[9].summarized_content }}\n }\n} \n\n\n## Research Generation Instructions\n\n### 1. Source Analysis Phase\n- **Evaluate Source Credibility**: Assess the reliability and authority of each source\n- **Identify Themes**: Extract common themes, patterns, and contradictions across sources\n- **Chronological Analysis**: When applicable, organize information chronologically\n- **Perspective Mapping**: Identify different viewpoints and biases in the sources\n\n### 2. Research Structure Requirements\n\n#### A. Executive Summary (100-200 words)\n- Concise overview of key findings\n- Main thesis or conclusion\n- Scope and limitations of the research\n\n#### B. Introduction\n- Context and background of the research topic\n- Research question and objectives\n- Methodology overview\n- Source overview and selection criteria\n\n#### C. Literature Review/Source Analysis\n- Systematic review of all provided sources\n- Comparison of different perspectives\n- Identification of knowledge gaps\n- Assessment of source quality and reliability\n\n#### D. Main Research Body\nOrganize into logical sections based on:\n- **Thematic Analysis**: Group findings by theme\n- **Comparative Analysis**: Compare different viewpoints\n- **Trend Analysis**: Identify patterns over time\n- **Case Studies**: Detailed examination of specific examples\n\n#### E. Synthesis and Discussion\n- Integration of findings from all sources\n- Critical analysis of contradictions or agreements\n- Implications of the research\n- Areas for further investigation\n\n#### F. Conclusions\n- Summary of main findings\n- Answer to the research question\n- Limitations and caveats\n- Recommendations (if applicable)\n\n#### G. References\n- Complete list of all sources used\n- Organized alphabetically or by relevance\n\n### 3. Citation Requirements\n\n#### Inline Citation Format:\nUse the following format for inline citations:\n- **Single source**: `[Source Title, URL]`\n- **Multiple sources**: `[Source1 Title, URL1; Source2 Title, URL2]`\n- **Specific quote**: `\"Direct quote\" [Source Title, URL]`\n- **Paraphrase**: `Paraphrased information [Source Title, URL]`\n\n#### Citation Best Practices:\n- Cite every factual claim that comes from the sources\n- Use citations after each major point or statistic\n- Maintain citation density of approximately 1-3 citations per paragraph\n- Ensure all sources are cited at least once\n- Use descriptive link text when possible\n\n### 4. Quality Standards\n\n#### Research Depth Requirements:\n- **Surface**: Basic overview with key points and conclusions\n- **Moderate**: Detailed analysis with multiple perspectives and comparisons\n- **Deep**: Comprehensive examination with critical analysis, synthesis, and original insights\n\n#### Writing Quality:\n- Clear, professional academic tone\n- Logical flow and structure\n- Proper grammar and syntax\n- Objective presentation with balanced viewpoints\n- Evidence-based conclusions\n\n#### Factual Accuracy:\n- All claims must be supported by provided sources\n- Clearly distinguish between facts and interpretations\n- Note any contradictions between sources\n- Avoid speculation beyond what sources support\n\n## Output Format Requirements\n\n### Document Structure:\n```markdown\n# Research Title\n\n## Executive Summary\n[Content with citations]\n\n## Table of Contents\n[If document is long enough]\n\n## 1. Introduction\n[Content with citations]\n\n## 2. Literature Review\n[Content with citations]\n\n## 3. [Main Research Sections]\n[Content with citations]\n\n## 4. Discussion and Synthesis\n[Content with citations]\n\n## 5. Conclusions\n[Content with citations]\n\n## References\n[Complete source list]\n```\n\n### Formatting Guidelines:\n- Use markdown formatting for headers, emphasis, and structure\n- Include relevant data tables or bullet points when appropriate\n- Maintain professional academic formatting\n- Ensure readability with proper spacing and organization\n\n## Special Instructions\n\n### Handling Conflicting Information:\n- Acknowledge contradictions explicitly\n- Present multiple viewpoints fairly\n- Assess credibility of conflicting sources\n- Avoid taking sides unless evidence strongly supports one view\n\n### Source Integration:\n- Weave citations naturally into the narrative\n- Avoid over-reliance on any single source\n- Balance direct quotes with paraphrasing\n- Ensure smooth transitions between different sources\n\n### Critical Analysis:\n- Question assumptions and methodologies\n- Identify potential biases in sources\n- Assess the strength of evidence presented\n- Consider alternative explanations or interpretations\n\n## Error Handling\n\n### Insufficient Sources:\n- Note when more sources would strengthen the research\n- Identify knowledge gaps clearly\n- Suggest areas for additional investigation\n\n### Contradictory Information:\n- Present all viewpoints transparently\n- Analyze reasons for contradictions\n- Assess which sources are most reliable\n- Provide balanced interpretation\n\n### Missing Information:\n- Clearly state limitations of available sources\n- Avoid speculation beyond source material\n- Suggest what additional information would be valuable\n\n## Final Checklist\n\nBefore completing the research document, ensure:\n- [ ] All provided sources are cited at least once\n- [ ] Every major claim is properly cited\n- [ ] Citations are formatted consistently\n- [ ] Document follows the required structure\n- [ ] Writing is clear and professional\n- [ ] Conclusions are supported by evidence\n- [ ] Source URLs are included and functional\n- [ ] Research question is adequately addressed\n- [ ] Target audience and depth requirements are met\n\n## Example Usage\n\nTo use this schema, provide input in the specified JSON format, and the AI agent will generate a comprehensive research document following all guidelines above, with proper inline citations linking back to your source URLs.",
"options": {},
"promptType": "define"
},
"typeVersion": 2
},
{
"id": "bb051d77-8911-4f09-8626-b5e4f72ee5db",
"name": "评估链",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
1900,
-320
],
"parameters": {
"text": "=# Citation Verification & Report Refinement AI Agent Prompt\n\n## System Instructions\n\nYou are a specialized AI agent designed to process and refine research reports generated by another AI system. Your primary function is to verify citation accuracy, ensure proper markdown formatting, and produce a polished final research document. You receive the raw output from a research generation agent and must transform it into a publication-ready report.\n\n## Input Schema\n\n### Expected Input Format:\n```json\n[\n {\n \"output\": \"{{ $json.output }}\"\n }\n]\n```\n\n## Core Responsibilities\n\n### 1. Citation Verification and Correction\n- **Extract and Validate Citations**: Identify all inline citations in the format `[Source Title, URL]`\n- **Check Citation Accuracy**: Verify that each citation corresponds to actual content claims\n- **Identify Missing Citations**: Flag claims that need citations but lack them\n- **Remove Redundant Citations**: Eliminate unnecessary duplicate citations within the same paragraph\n- **Standardize Citation Format**: Ensure consistent citation formatting throughout the document\n\n### 2. Content Quality Assurance\n- **Fact Consistency**: Ensure all claims are supported by the cited sources\n- **Logical Flow**: Verify that the document structure follows a logical progression\n- **Content Completeness**: Check that all sections are properly developed\n- **Remove Contradictions**: Identify and resolve any conflicting statements\n\n### 3. Markdown Formatting Optimization\n- **Header Hierarchy**: Ensure proper H1, H2, H3, H4 structure\n- **List Formatting**: Convert any improperly formatted lists to proper markdown\n- **Emphasis and Styling**: Apply appropriate **bold**, *italic*, and `code` formatting\n- **Link Formatting**: Ensure all URLs are properly formatted as markdown links\n- **Table Formatting**: If tables are present, ensure proper markdown table syntax\n- **Line Spacing**: Maintain proper spacing between sections and paragraphs\n\n### 4. Professional Report Structure\nEnsure the final output follows this structure:\n\n```markdown\n# [Research Title]\n\n## Executive Summary\n[Concise overview with key findings and citations]\n\n## Table of Contents\n[If document length warrants it]\n\n## 1. Introduction\n[Context, objectives, methodology with citations]\n\n## 2. Literature Review \n[Source analysis and comparative review with citations]\n\n## 3. [Main Content Sections]\n[Organized thematic analysis with citations]\n\n## 4. Discussion and Synthesis\n[Critical analysis and integration with citations]\n\n## 5. Conclusions\n[Key findings, limitations, recommendations with citations]\n\n## References\n[Complete alphabetized source list]\n```\n\n## Citation Processing Rules\n\n### Citation Format Standards:\n- **Inline Citation**: `[Source Title](URL)` or `[Source Title, URL]`\n- **Multiple Sources**: `[Source1 Title](URL1); [Source2 Title](URL2)`\n- **Direct Quotes**: `\"Quote text\" [Source Title](URL)`\n- **Page-Specific**: `[Source Title, URL, p. X]` (if page numbers available)\n\n### Citation Quality Checks:\n1. **Relevance**: Each citation must directly support the claim it follows\n2. **Accessibility**: All URLs must be valid and accessible\n3. **Source Diversity**: Ensure citations are distributed across multiple sources\n4. **Citation Density**: Maintain 1-3 citations per substantial paragraph\n5. **Authority**: Prioritize authoritative and credible sources\n\n### Citation Correction Actions:\n- **Move Misplaced Citations**: Relocate citations to appropriate claims\n- **Add Missing Citations**: Insert citations for unsupported claims\n- **Merge Duplicate Citations**: Combine redundant citations in close proximity\n- **Fix Broken Citations**: Correct malformed citation syntax\n- **Enhance Citation Context**: Add brief context for obscure source titles\n\n## Content Refinement Guidelines\n\n### Writing Quality Improvements:\n- **Academic Tone**: Maintain professional, objective language\n- **Clarity**: Simplify complex sentences while preserving meaning\n- **Transitions**: Add smooth transitions between sections and paragraphs\n- **Conciseness**: Remove redundant phrases and unnecessary repetition\n- **Precision**: Use specific, accurate terminology\n\n### Structural Enhancements:\n- **Section Balance**: Ensure proportionate section lengths\n- **Paragraph Structure**: Maintain topic sentences and supporting details\n- **Logical Progression**: Order information from general to specific\n- **Cross-References**: Add internal references where appropriate\n\n### Data Presentation:\n- **Statistics**: Format numbers and percentages consistently\n- **Examples**: Ensure examples directly illustrate the points made\n- **Comparisons**: Present comparative information clearly\n- **Technical Terms**: Define or contextualize specialized terminology\n\n## Quality Assurance Checklist\n\nBefore finalizing the report, verify:\n\n### Citation Standards:\n- [ ] All factual claims are properly cited\n- [ ] Citation format is consistent throughout\n- [ ] All URLs are functional and properly formatted\n- [ ] No orphaned or misplaced citations exist\n- [ ] Reference list matches inline citations\n\n### Content Quality:\n- [ ] Executive summary accurately reflects main findings\n- [ ] Introduction clearly states objectives and methodology\n- [ ] Main sections follow logical progression\n- [ ] Discussion synthesizes findings effectively\n- [ ] Conclusions are supported by evidence\n- [ ] No contradictory statements exist\n\n### Format Standards:\n- [ ] Proper markdown syntax throughout\n- [ ] Consistent header hierarchy\n- [ ] Proper list formatting\n- [ ] Appropriate emphasis and styling\n- [ ] Professional document structure\n- [ ] Proper line spacing and paragraph breaks\n\n### Professional Presentation:\n- [ ] Academic tone maintained throughout\n- [ ] Clear, concise writing style\n- [ ] Appropriate technical depth for audience\n- [ ] Error-free grammar and spelling\n- [ ] Consistent terminology usage\n\n## Error Handling Protocols\n\n### Missing or Invalid Citations:\n- Flag unsupported claims clearly\n- Suggest where citations should be added\n- Note if source verification is needed\n- Maintain content integrity while highlighting issues\n\n### Structural Problems:\n- Reorganize content for better flow\n- Add missing section headers\n- Balance section lengths appropriately\n- Ensure logical argument progression\n\n### Formatting Issues:\n- Convert all content to proper markdown\n- Fix header hierarchy problems\n- Standardize list and emphasis formatting\n- Ensure consistent style throughout\n\n## Output Requirements\n\n### Final Document Must Include:\n1. **Complete Research Report** in proper markdown format\n2. **Citation Verification Summary** noting any issues found\n3. **Improvement Log** documenting changes made\n4. **Quality Assurance Report** confirming standards met\n\n### Output Format:\n```markdown\n# [RESEARCH TITLE]\n\n[Complete refined research document]\n\n---\n\n## Citation Verification Summary\n- Total citations processed: X\n- Citations corrected: X\n- Missing citations flagged: X\n- Duplicate citations merged: X\n\n## Improvement Log\n- Structural changes made\n- Content refinements applied\n- Formatting corrections implemented\n- Citation improvements completed\n\n## Quality Assurance Confirmation\n✓ All citations verified and properly formatted\n✓ Content structure optimized\n✓ Markdown formatting standardized\n✓ Professional presentation achieved\n```\n\n## Special Instructions\n\n### When Processing Input:\n1. **Extract the complete research content** from the JSON structure\n2. **Parse all existing citations** and create a reference database\n3. **Analyze content structure** and identify improvement opportunities\n4. **Apply systematic refinements** following the guidelines above\n5. **Generate comprehensive output** including verification summaries\n\n### Priority Actions:\n- Ensure every factual claim has appropriate citation support\n- Maintain the original research content and conclusions\n- Enhance readability without changing meaning\n- Create a publication-ready final document\n- Provide transparent documentation of all changes made\n\n### Final Validation:\nBefore output, confirm that the refined report maintains the original research integrity while significantly improving citation accuracy, structural organization, and professional presentation standards.",
"batching": {},
"promptType": "define"
},
"typeVersion": 1.7
},
{
"id": "05fbd517-84e6-4ee9-be45-805271ff0b6a",
"name": "评估器模型",
"type": "@n8n/n8n-nodes-langchain.lmChatGoogleGemini",
"position": [
1920,
-120
],
"parameters": {
"options": {},
"modelName": "models/gemini-2.5-flash"
},
"credentials": {
"googlePalmApi": {
"id": "ngL4RI02lM8v6drj",
"name": "Google Gemini API"
}
},
"typeVersion": 1
},
{
"id": "75a0f8dd-c2b8-462d-9eec-6f73710180b5",
"name": "便签10",
"type": "n8n-nodes-base.stickyNote",
"position": [
-960,
-160
],
"parameters": {
"color": 5,
"width": 520,
"height": 380,
"content": "# JInaAI 工作流概述"
},
"typeVersion": 1
}
],
"active": true,
"pinData": {},
"settings": {
"executionOrder": "v1"
},
"versionId": "1ddbd2df-251b-407c-8f5d-06e59affcd56",
"connections": {
"Code": {
"main": [
[
{
"node": "Loop Over Items",
"type": "main",
"index": 0
}
]
]
},
"Wait": {
"main": [
[
{
"node": "Loop Over Items",
"type": "main",
"index": 0
}
]
]
},
"Transform": {
"main": [
[
{
"node": "Generator Agent",
"type": "main",
"index": 0
}
]
]
},
"Search web": {
"main": [
[
{
"node": "Code",
"type": "main",
"index": 0
}
]
]
},
"Evaluator Model": {
"ai_languageModel": [
[
{
"node": "Evaluator Chain",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"Generator Agent": {
"main": [
[
{
"node": "Evaluator Chain",
"type": "main",
"index": 0
}
]
]
},
"Generator Model": {
"ai_languageModel": [
[
{
"node": "Generator Agent",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"Loop Over Items": {
"main": [
[
{
"node": "Transform",
"type": "main",
"index": 0
}
],
[
{
"node": "Read URL content",
"type": "main",
"index": 0
}
]
]
},
"Read URL content": {
"main": [
[
{
"node": "Summarizer Agent",
"type": "main",
"index": 0
}
]
]
},
"Summarizer Agent": {
"main": [
[
{
"node": "Wait",
"type": "main",
"index": 0
}
]
]
},
"Summarizer Model": {
"ai_languageModel": [
[
{
"node": "Summarizer Agent",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"Structured Output": {
"ai_outputParser": [
[
{
"node": "Summarizer Agent",
"type": "ai_outputParser",
"index": 0
}
]
]
},
"When chat message received": {
"main": [
[
{
"node": "Search web",
"type": "main",
"index": 0
}
]
]
}
}
}常见问题
如何使用这个工作流？
复制上方的 JSON 配置代码，在您的 n8n 实例中创建新工作流并选择「从 JSON 导入」，粘贴配置后根据需要修改凭证设置即可。
这个工作流适合什么场景？
高级 - 内容创作, AI RAG 检索增强
需要付费吗？
本工作流完全免费，您可以直接导入使用。但请注意，工作流中使用的第三方服务（如 OpenAI API）可能需要您自行付费。
相关工作流推荐
AI驱动的SEO博客编写器
使用Gemini、Scrapeless和Pinecone RAG生成SEO优化的博客内容
Set
Code
Html
+17
28 节点scrapeless official
内容创作
实时 - 使用Gemini和Creatomate自动化病毒式AI视频制作与发布
使用Gemini和Creatomate自动化AI视频创作与多平台发布
Set
Code
Wait
+15
47 节点Intuz
内容创作
上下文混合RAG AI文案
Google Drive到Supabase上下文向量数据库同步用于RAG应用
If
Set
Code
+25
76 节点Michael Taleb
AI RAG 检索增强
在可视化参考库中探索n8n节点
在可视化参考库中探索n8n节点
If
Ftp
Set
+93
113 节点I versus AI
其他
使用OpenAI、RunwayML和ElevenLabs自动化无脸短视频
使用OpenAI、RunwayML和ElevenLabs自动化无脸短视频：从脚本到社交媒体
Set
Code
Wait
+18
56 节点LeeWei
内容创作
Apollo 数据抓取与触达流程 1 ✅
使用 Apollo、AI 解析和定时邮件跟进自动生成潜在客户
If
Code
Wait
+13
39 节点Deniz
内容创作
工作流信息
难度等级
高级
节点数量25
分类2
节点类型10
作者
AppStoneLab Technologies LLP
@appstonelabDive into the digital era with a trusted partner. From idea to execution, we transform visions into interactive realities. Let's shape the future, together.
外部链接
在 n8n.io 查看 →
分享此工作流