Asistente de investigación IA usando Jina y Gemini

• •Clave de API de Google Gemini

{
  "id": "v9dRFBMYVwnQUhii",
  "meta": {
    "instanceId": "d06afcec45b6881aa3c2bd3e34579ea44e46952db235308e9371e2bb743e4b43",
    "templateCredsSetupCompleted": true
  },
  "name": "AI Research Assistant with Jina & Gemini",
  "tags": [],
  "nodes": [
    {
      "id": "43de71f4-db4d-41aa-8294-abf2e0e984c8",
      "name": "Buscar en la web",
      "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": "Cuando se recibe un mensaje en el chat",
      "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": "Iterar sobre elementos",
      "type": "n8n-nodes-base.splitInBatches",
      "position": [
        700,
        40
      ],
      "parameters": {
        "options": {
          "reset": false
        }
      },
      "typeVersion": 3
    },
    {
      "id": "0a8274f0-7ef4-4305-a3e7-81b34cd07ddc",
      "name": "Código",
      "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": "Leer contenido de URL",
      "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": "Esperar",
      "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": "Salida estructurada",
      "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": "Transformar",
      "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": "Nota adhesiva",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -420,
        -160
      ],
      "parameters": {
        "width": 340,
        "height": 380,
        "content": "## User query\n"
      },
      "typeVersion": 1
    },
    {
      "id": "0bb609ef-222d-4cc6-8ffd-0145cea615f3",
      "name": "Nota adhesiva1",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -60,
        -160
      ],
      "parameters": {
        "color": 7,
        "width": 260,
        "height": 380,
        "content": "## Search web\n\n**Purpose:** Web search using Jina AI based on chat input.\n**Input:** Chat message.\n**Note:** Requires Jina AI API key."
      },
      "typeVersion": 1
    },
    {
      "id": "01ee087c-e26c-4b8f-9f3f-9598d0cd83d3",
      "name": "Nota adhesiva2",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        220,
        -160
      ],
      "parameters": {
        "color": 2,
        "width": 320,
        "height": 380,
        "content": "## Code (URL Transform)\n\n**Purpose:** Custom JavaScript code to process/transform URLs, likely preparing them for the loop.\n**Input:** Raw URLs from \"Search web\".\n**Output:** Formatted URLs suitable for \"Loop over Items\"."
      },
      "typeVersion": 1
    },
    {
      "id": "76f010e5-fc84-4066-a39a-7dd870ff9345",
      "name": "Nota adhesiva3",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        560,
        -160
      ],
      "parameters": {
        "color": 6,
        "width": 380,
        "height": 380,
        "content": "## Loop over all URLs\n\n**Purpose:** Iterates over each URL provided by the \"Code\" node.\n**Function:** Processes each URL individually for content gathering and summarization.\n**Note:** Essential for handling multiple search results."
      },
      "typeVersion": 1
    },
    {
      "id": "6cdc7a5f-b488-4b72-a0da-9351c7478ea1",
      "name": "Nota adhesiva4",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        960,
        100
      ],
      "parameters": {
        "color": 3,
        "width": 260,
        "height": 480,
        "content": "## Read URL Content\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n**Purpose:** Fetches the full content of each URL provided by the loop.\n**Input:** Individual URLs from \"Loop over Items\".\n**Output:** Raw web page content.\n**Note:** Essential for summarization."
      },
      "typeVersion": 1
    },
    {
      "id": "0d3c40cd-733b-4508-adc6-e34171e48ef1",
      "name": "Nota adhesiva5",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        1240,
        100
      ],
      "parameters": {
        "color": 5,
        "width": 600,
        "height": 480,
        "content": "## Summarize Gathered content\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n**Purpose:** AI agent to summarize web page content.\n**Input:** Web content & user query.\n**Output:** Summarized content.\n**Note:** Uses \"Summarizer Model\"."
      },
      "typeVersion": 1
    },
    {
      "id": "30e03e88-16a5-4a77-bf19-74488042d88f",
      "name": "Nota adhesiva6",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        1860,
        100
      ],
      "parameters": {
        "color": 4,
        "width": 340,
        "height": 480,
        "content": "## Avoid rate limits\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n**Purpose:** Introduces a pause to prevent hitting API rate limits.\n**Configuration:** Adjust delay time (e.g., 1000ms) based on API limits.\n**Tip:** Crucial for stable, long-running processes."
      },
      "typeVersion": 1
    },
    {
      "id": "52fd4d1c-63b6-44f0-b43f-30f092eadf7f",
      "name": "Nota adhesiva7",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        960,
        -480
      ],
      "parameters": {
        "width": 300,
        "height": 480,
        "content": "## Transform (Summaries Aggregator)\n\n**Purpose:** Python code to consolidate all individual summarized outputs into a single list.\n**Input:** Multiple summaries.\n**Output:** One item with `{\"output\": [all_summaries]}`.\n**Key:** Prepares data for research report generation."
      },
      "typeVersion": 1
    },
    {
      "id": "1572a6bd-ed9b-4461-9167-4b1d98bfc878",
      "name": "Nota adhesiva8",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        1280,
        -480
      ],
      "parameters": {
        "color": 7,
        "width": 480,
        "height": 480,
        "content": "## Generate research Report\n\n**Purpose:** AI Agent to orchestrate research report generation.\n**Input:** Summarized data from \"Transform summarized data\".\n**Function:** Guides the \"Generator Model\" to create the report."
      },
      "typeVersion": 1
    },
    {
      "id": "f35efe92-7181-4283-8b93-858b807b1b40",
      "name": "Nota adhesiva9",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        1780,
        -480
      ],
      "parameters": {
        "width": 500,
        "height": 480,
        "content": "## Evaluate and Generate Final Report\n\n**Purpose:** AI Chain to evaluate and refine the generated report.\n**Input:** Research report from \"Generator Agent\".\n**Function:** Uses \"Evaluator Model\" to assess and improve output quality."
      },
      "typeVersion": 1
    },
    {
      "id": "d6d4b224-dd71-4bf7-be6f-b804002d48ac",
      "name": "Agente de resumen",
      "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": "Modelo de resumen",
      "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": "Modelo generador",
      "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": "Agente generador",
      "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": "Cadena de evaluación",
      "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": "Modelo de evaluación",
      "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": "Nota adhesiva10",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -960,
        -160
      ],
      "parameters": {
        "color": 5,
        "width": 520,
        "height": 380,
        "content": "# JInaAI Workflow Overview\n\nThis workflow intelligently processes incoming chat messages by performing a web search using Jina AI and then leverages Gemini 2.5 Flash to generate and evaluate responses.\n\n**How to Use:**\n* Send a chat message to the configured webhook for the \"When chat message received\" node.\n* The workflow will automatically search the web and generate an AI-powered response.\n\n**Expected Output:**\n* An AI-generated response, potentially incorporating information from web searches."
      },
      "typeVersion": 1
    }
  ],
  "active": true,
  "pinData": {},
  "settings": {
    "executionOrder": "v1"
  },
  "versionId": "1ddbd2df-251b-407c-8f5d-06e59affcd56",
  "connections": {
    "0a8274f0-7ef4-4305-a3e7-81b34cd07ddc": {
      "main": [
        [
          {
            "node": "0c414f7e-d927-4a6b-a45c-412676d99986",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "e282b54c-4230-449b-ab7c-96f0566e81f2": {
      "main": [
        [
          {
            "node": "0c414f7e-d927-4a6b-a45c-412676d99986",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "ef58119f-ab0e-4c99-a83d-d2a0dbc0ba17": {
      "main": [
        [
          {
            "node": "bc46ca58-24a6-4942-86e2-1c4c4df46c45",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "43de71f4-db4d-41aa-8294-abf2e0e984c8": {
      "main": [
        [
          {
            "node": "0a8274f0-7ef4-4305-a3e7-81b34cd07ddc",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "05fbd517-84e6-4ee9-be45-805271ff0b6a": {
      "ai_languageModel": [
        [
          {
            "node": "bb051d77-8911-4f09-8626-b5e4f72ee5db",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "bc46ca58-24a6-4942-86e2-1c4c4df46c45": {
      "main": [
        [
          {
            "node": "bb051d77-8911-4f09-8626-b5e4f72ee5db",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "4d26eb57-9e13-45e9-b365-c531189b0030": {
      "ai_languageModel": [
        [
          {
            "node": "bc46ca58-24a6-4942-86e2-1c4c4df46c45",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "0c414f7e-d927-4a6b-a45c-412676d99986": {
      "main": [
        [
          {
            "node": "ef58119f-ab0e-4c99-a83d-d2a0dbc0ba17",
            "type": "main",
            "index": 0
          }
        ],
        [
          {
            "node": "540cae13-9d70-47b5-9be4-902abc7d6391",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "540cae13-9d70-47b5-9be4-902abc7d6391": {
      "main": [
        [
          {
            "node": "d6d4b224-dd71-4bf7-be6f-b804002d48ac",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "d6d4b224-dd71-4bf7-be6f-b804002d48ac": {
      "main": [
        [
          {
            "node": "e282b54c-4230-449b-ab7c-96f0566e81f2",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "ef2cf6a3-3ea5-4874-898f-25ebc56997f1": {
      "ai_languageModel": [
        [
          {
            "node": "d6d4b224-dd71-4bf7-be6f-b804002d48ac",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "37c6d91d-526e-4c9a-a1f3-4bbba63740bd": {
      "ai_outputParser": [
        [
          {
            "node": "d6d4b224-dd71-4bf7-be6f-b804002d48ac",
            "type": "ai_outputParser",
            "index": 0
          }
        ]
      ]
    },
    "8126e0ee-8472-4f1f-8921-1e2aded6682e": {
      "main": [
        [
          {
            "node": "43de71f4-db4d-41aa-8294-abf2e0e984c8",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  }
}