Research & Content Gathering App

A powerful, full-featured research and content gathering application with live parameter adjustment capabilities. Built with Python and FastMCP, this app provides an intuitive web interface for conducting research, gathering content, and building structured outputs.

Features

Core Capabilities

• Full Research Engine: Advanced web scraping and content gathering

• Live Parameter Adjustment: Modify research parameters in real-time while tasks are running

• Content Analysis: Automatic relevance scoring, summarization, and topic extraction

• Content Building: Generate articles, reports, summaries, and presentations from research

• Real-time Progress Tracking: Monitor research progress with live updates

• Multiple Output Formats: Export as Markdown, HTML, JSON, or plain text

• Configuration Management: Save and load research configurations

Multi-Platform Asset Generation

Generate professional assets for multiple platforms:

• YouTube: Video scripts, thumbnail templates, descriptions, SEO tags

• Gumroad: Product listings, pricing strategies, marketing copy

• Etsy: Optimized product listings, tags, descriptions, SEO

• Websites: Landing pages, blog posts, SEO elements

• Game Development: Design documents, mechanics, story structures

• Canva: Template specifications, design guides, color palettes

• Multi-Platform Packages: Generate complete asset sets for product launches

Web Interface Features

• Beautiful, responsive UI with gradient design

• Real-time parameter sliders (depth, max results, relevance threshold, detail level)

• Interactive tabs for Results, Statistics, Logs, and Content Builder

• Live progress bar and status indicators

• Activity logging with timestamps

• Statistics dashboard with quality metrics

MCP Tools (23+ Available)

Research Tools:

1. start_research - Start research with customizable parameters

2. get_research_status - Check current task status and progress

3. pause_research - Pause running research tasks

4. resume_research - Resume paused tasks

5. stop_research - Stop current research

6. update_parameters - Live parameter updates while running

7. get_results - Retrieve results in various formats

8. export_results - Export to files

9. build_content - Generate structured content (articles, reports, etc.)

10. analyze_results - Get detailed analysis and insights

11. get_statistics - View session statistics

12. search_results - Search within collected results

13. save_configuration - Save research settings

14. load_configuration - Load saved settings

Asset Generation Tools: 15. generate_youtube_assets - Create video scripts, thumbnails, descriptions 16. generate_gumroad_listing - Generate product listings for Gumroad 17. generate_etsy_listing - Create optimized Etsy product listings 18. generate_web_assets - Build landing pages and blog templates 19. generate_game_assets - Create game design documents 20. generate_canva_specs - Generate Canva template specifications 21. list_platforms - Show all supported platforms 22. generate_multi_platform_package - Create complete asset packages

Prompts: 23. research_prompt - Generate research prompts 24. content_prompt - Generate content creation prompts

Installation

Prerequisites

• Python 3.13+

• uv package manager

Setup

1. Clone the repository:

2. Install Python version and dependencies:

Usage

Starting the Server

Run the server on port 3000:

The server will start and display:

Using the Web Interface

1. Open your browser to http://127.0.0.1:3000/static/index.html

2. Enter your research topic in the query field

3. Adjust parameters using the live sliders:

   • Depth Level (1-10): How deep to research

   • Max Results (5-100): Maximum number of results

   • Relevance Threshold (0.0-1.0): Minimum quality score

   • Content Detail (1-10): Amount of content to extract

4. Select source type and output format

5. Click "Start Research" to begin

6. Adjust parameters in real-time while research runs

7. Use Pause/Resume/Stop controls as needed

8. View results in the Results tab

9. Check Statistics for insights

10. Build content from research in the Content Builder tab

Using MCP Tools

You can also interact with the server programmatically through MCP tools:

Generating Assets for Different Platforms

YouTube Content Creation:

Gumroad Product Launch:

Etsy Shop:

Website Development:

Game Development:

Testing with MCP Inspector

1. Install Node.js (^22.7.5)

2. Run the inspector:

3. Open http://localhost:6274

4. Configure:

   • Transport Type: Streamable HTTP

   • URL: http://127.0.0.1:3000/mcp

Project Structure

Research Parameters

Query Options

• query: Research topic (required)

• source_type: all, academic, news, technical, social

• depth: Research depth level (1-10)

• max_results: Maximum results to gather (5-100)

• relevance_threshold: Minimum quality score (0.0-1.0)

• detail_level: Content extraction detail (1-10)

• output_format: markdown, html, json, text

• summary_length: brief, moderate, detailed

Content Building Options

Build different types of content from research:

• Article: Structured article with introduction and conclusions

• Report: Formal research report with executive summary

• Summary: Brief executive summary of findings

• Presentation: Slide-formatted content

Tone Options

• Professional

• Academic

• Casual

• Technical

Statistics & Analytics

The app tracks comprehensive statistics:

• Total queries performed

• Total results gathered

• Average relevance score

• Processing time

• Sources analyzed

• Error count

• Quality score (composite metric)

• Topic distribution

• Source type distribution

Live Parameter Adjustment

One of the key features is the ability to adjust parameters while research is running:

1. Start a research task

2. Move sliders in the sidebar

3. Parameters update automatically

4. Research adapts to new settings in real-time

This allows you to:

• Increase depth if initial results are shallow

• Raise/lower relevance threshold to filter results

• Adjust max results to gather more or less data

• Change detail level for more/less content

Export & Sharing

Export research results in multiple formats:

• Markdown: Perfect for documentation

• HTML: Ready for web publishing

• JSON: For programmatic processing

• Text: Simple plain text format

Save configurations for reuse:

Development

Adding Custom Search Providers

Edit research_engine.py to add real search integrations:

Extending Content Builders

Add new content types in research_engine.py:

Then register in main.py:

Custom Analysis Tools

Add specialized analysis in the research engine:

Architecture

Research Engine

• Asynchronous research execution

• Parallel content fetching

• Real-time parameter updates

• Pause/resume capability

• Progress tracking

Content Processing

1. Web search and URL discovery

2. Content fetching (with timeout handling)

3. HTML parsing and text extraction

4. Relevance scoring

5. Summarization

6. Topic analysis

7. Quality metrics

MCP Integration

• FastMCP server with streamable HTTP transport

• 15+ tools for comprehensive research control

• Resource serving for web interface

• Prompt templates for research and content generation

Performance

The app is designed for efficiency:

• Asynchronous I/O for parallel processing

• Smart throttling to avoid rate limits

• Configurable timeouts

• Progress tracking without blocking

• Memory-efficient result storage

Limitations & Future Enhancements

Current limitations:

• Web search uses placeholder (requires API integration)

• Basic text extraction (can be enhanced with ML models)

• Simple relevance scoring (can use embeddings/semantic search)

Planned enhancements:

• Integration with real search APIs

• Advanced NLP for better summarization

• Semantic similarity scoring

• Citation network analysis

• Multi-language support

• Database storage for large result sets

• WebSocket for true real-time updates

• Collaborative research sessions

Troubleshooting

Port already in use

Dependencies issues

Web interface not loading

• Check that static/index.html exists

• Ensure server is running on port 3000

• Try accessing http://127.0.0.1:3000 directly

License

This project is open source and available under the MIT License.

Contributing

Contributions are welcome! Please:

1. Fork the repository

2. Create a feature branch

3. Add tests for new features

4. Submit a pull request

Support

For issues, questions, or suggestions:

• Open an issue on GitHub

• Check existing documentation

• Review MCP documentation at https://modelcontextprotocol.io

Credits

Built with:

• FastMCP - MCP server framework

• aiohttp - Async HTTP client

• BeautifulSoup - HTML parsing

• Python Markdown - Markdown processing

Happy Researching! 🔬📚✨