"""
Intelligent prompts for RivalSearchMCP.
Guides users on effective tool usage patterns.
"""
from fastmcp import FastMCP
def register_prompts(mcp: FastMCP):
"""Register all prompts for RivalSearchMCP."""
@mcp.prompt
def comprehensive_research(topic: str, depth: str = "comprehensive") -> list:
"""
Generate a comprehensive research workflow using the research_agent.
This prompt guides the AI to use the research_agent tool effectively,
which autonomously searches web, social media, news, GitHub, and academic
sources to create a detailed research report.
Args:
topic: The research topic or question
depth: Research depth (basic, comprehensive, expert)
"""
return [
{
"role": "user",
"content": f"""I need you to conduct comprehensive research on: {topic}
Use the research_agent tool with these parameters:
- topic: "{topic}"
- research_depth: "{depth}"
- max_sources: 15
- enable_ai_insights: true
- include_trends: false
- include_website_analysis: true
The research_agent will autonomously:
1. Search across DuckDuckGo, Yahoo, and Wikipedia
2. Explore social media discussions (Reddit, Hacker News, etc.)
3. Find relevant news articles
4. Search GitHub repositories
5. Analyze websites and documents
6. Generate a 4000+ character professional report
After receiving the research report, analyze it and provide:
- Key findings summary
- Important insights
- Recommendations based on the research
- Areas that need further investigation"""
}
]
@mcp.prompt
def multi_source_search(query: str, include_social: bool = True, include_news: bool = True) -> list:
"""
Search across multiple sources (web, social, news) and synthesize results.
Guides the workflow: web_search → social_search → news_aggregation → synthesis
Args:
query: Search query
include_social: Whether to include social media search
include_news: Whether to include news aggregation
"""
steps = [f"""Search for information about: {query}
Follow this research workflow:
1. First, use web_search to find general information:
- query: "{query}"
- num_results: 5"""]
if include_social:
steps.append("""
2. Then, use social_search to find community discussions:
- platforms: ["reddit", "hackernews", "devto"]
- max_results_per_platform: 5""")
if include_news:
steps.append("""
3. Next, use news_aggregation to find recent news:
- max_results: 5""")
steps.append("""
Finally, analyze all the information gathered and provide:
- Summary of web search findings
- Key insights from social discussions
- Recent developments from news
- Comprehensive synthesis of all sources""")
return [{"role": "user", "content": "\n".join(steps)}]
@mcp.prompt
def deep_content_analysis(url: str, extract_documents: bool = False) -> list:
"""
Perform deep analysis of a website and its content.
Guides workflow: map_website → content_operations → document_analysis
Args:
url: Website URL to analyze
extract_documents: Whether to analyze linked documents (PDFs, etc.)
"""
prompt = f"""Perform a deep analysis of this website: {url}
Follow this workflow:
1. Use map_website to explore the site structure:
- url: "{url}"
- mode: "research"
- max_pages: 10
2. For interesting pages found, use content_operations to retrieve full content:
- operation: "retrieve"
- extraction_method: "markdown"
3. Use content_operations to extract all links:
- operation: "extract"
- link_type: "all"
"""
if extract_documents:
prompt += """
4. For any PDF or document links found, use document_analysis:
- Extract text with OCR support
- Analyze document content
"""
prompt += """
Finally, provide:
- Website structure overview
- Key content themes
- Important pages and their purposes
- Document summaries (if applicable)
- Comprehensive site analysis"""
return [{"role": "user", "content": prompt}]
@mcp.prompt
def academic_literature_review(research_question: str, max_papers: int = 10) -> list:
"""
Conduct an academic literature review with document analysis.
Guides workflow: scientific_research → document_analysis → synthesis
Args:
research_question: The research question or topic
max_papers: Maximum number of papers to review
"""
return [
{
"role": "user",
"content": f"""Conduct an academic literature review on: {research_question}
Follow this research workflow:
1. Use scientific_research to find relevant papers:
- operation: "academic_search"
- query: "{research_question}"
- max_results: {max_papers}
- sources: ["arxiv", "semantic_scholar"]
2. For papers with PDF links, use document_analysis to extract full text:
- Extract text from PDFs
- Use OCR for scanned papers if needed
3. Analyze and synthesize the findings:
- Identify common themes and patterns
- Note contradictions or debates
- Highlight key methodologies
- Summarize main conclusions
Provide a comprehensive literature review including:
- Overview of the research landscape
- Key papers and their contributions
- Common methodologies and approaches
- Research gaps and future directions
- Citations and references"""
}
]
