# Advanced Research System (Based on Anthropic's Paper)
[](https://discord.gg/EamjgSaEQf) [](https://www.youtube.com/@kyegomez3242) [](https://www.linkedin.com/in/kye-g-38759a207/) [](https://x.com/kyegomezb)
[](https://badge.fury.io/py/advanced_research)
[](https://www.python.org/downloads/release/python-3100/)
[](https://opensource.org/licenses/MIT)
An enhanced implementation of the orchestrator-worker pattern from Anthropic's paper, ["How we built our multi-agent research system"](https://www.anthropic.com/engineering/built-multi-agent-research-system), built on top of the bleeding-edge multi-agent framework `swarms`. Our implementation of this advanced research system leverages parallel execution, LLM-as-judge evaluation, and professional report generation with export capabilities.
## Installation
```bash
pip3 install -U advanced-research
# uv pip install -U advanced-research
```
## Environment Variables
```bash
# Exa Search API Key (Required for web search functionality)
EXA_API_KEY="your_exa_api_key_here"
# Anthropic API Key (For Claude models)
ANTHROPIC_API_KEY="your_anthropic_api_key_here"
# OpenAI API Key (For GPT models)
OPENAI_API_KEY="your_openai_api_key_here"
# Worker Agent Configuration
WORKER_MODEL_NAME="gpt-4.1"
WORKER_MAX_TOKENS=8000
# Exa Search Configuration
EXA_SEARCH_NUM_RESULTS=2
EXA_SEARCH_MAX_CHARACTERS=100
```
**Note:** At minimum, you need `EXA_API_KEY` for web search functionality. For LLM functionality, you need either `ANTHROPIC_API_KEY` or `OPENAI_API_KEY`.
## Quick Start
### Basic Usage
```python
from advanced_research import AdvancedResearch
# Initialize the research system
research_system = AdvancedResearch(
name="AI Research Team",
description="Specialized AI research system",
max_loops=1,
)
# Run research and get results
result = research_system.run(
"What are the latest developments in quantum computing?"
)
print(result)
```
### With Export Functionality
```python
from advanced_research import AdvancedResearch
# Initialize with export enabled
research_system = AdvancedResearch(
name="Quantum Computing Research",
description="Research team focused on quantum computing advances",
max_loops=1,
export_on=True, # Enable JSON export
)
# Run research - will automatically export to JSON file
research_system.run(
"What are the latest developments in quantum computing?"
)
# Results will be saved to a timestamped JSON file
```
### Advanced Configuration
```python
from advanced_research import AdvancedResearch
# Initialize with custom settings
research_system = AdvancedResearch(
name="Medical Research Team",
description="Specialized medical research system",
director_model_name="claude-3-5-sonnet-20250115", # Use latest Claude model
worker_model_name="claude-3-5-sonnet-20250115",
director_max_tokens=10000,
max_loops=2, # Multiple research iterations
output_type="all", # Include full conversation history
export_on=True,
)
# Run research with image input (if applicable)
result = research_system.run(
"What are the most effective treatments for Type 2 diabetes?",
img=None # Optional image input
)
```
### Batch Processing Multiple Queries
```python
from advanced_research import AdvancedResearch
# Initialize the system
research_system = AdvancedResearch(
name="Batch Research System",
max_loops=1,
export_on=True,
)
# Process multiple research tasks
tasks = [
"Latest advances in renewable energy storage",
"Current state of autonomous vehicle technology",
"Recent breakthroughs in cancer immunotherapy"
]
# Run batch processing
research_system.batched_run(tasks)
```
### Using Different Output Formats
```python
from advanced_research import AdvancedResearch
# Initialize with specific output type
research_system = AdvancedResearch(
name="Research System",
output_type="json", # Options: "all", "json", "markdown"
export_on=False, # Get results directly instead of exporting
)
# Run research and get formatted output
result = research_system.run(
"What are the key challenges in AGI development?"
)
# Check available output methods
available_formats = research_system.get_output_methods()
print(f"Available output formats: {available_formats}")
```
## Quick Reference
| Task | Code | Documentation |
|------|------|---------------|
| **Basic Research** | `AdvancedResearch().run("query")` | [Basic Usage →](DOCS.md#basic-research-setup) |
| **Export Results** | `AdvancedResearch(export_on=True)` | [Export Config →](DOCS.md#configuration-examples) |
| **Batch Processing** | `system.batched_run([queries])` | [Batch Processing →](DOCS.md#batch-processing-setup) |
| **Custom Models** | `AdvancedResearch(director_model_name="model")` | [Advanced Config →](DOCS.md#advanced-multi-loop-research) |
| **Output Formats** | `AdvancedResearch(output_type="json")` | [Output Types →](DOCS.md#types-and-enums) |
## Examples
Ready-to-run examples demonstrating all features of the Advanced Research system:
| Example | Description | File |
|---------|-------------|------|
| **Basic Usage** | Simple research with minimal configuration | [`examples/basic_usage.py`](examples/basic_usage.py) |
| **Export Functionality** | Save research results to JSON files | [`examples/export_example.py`](examples/export_example.py) |
| **Advanced Configuration** | Custom models, tokens, and multiple loops | [`examples/advanced_config.py`](examples/advanced_config.py) |
| **Custom Models** | Different AI model configurations | [`examples/custom_models.py`](examples/custom_models.py) |
| **Output Formats** | JSON, markdown, and conversation history | [`examples/output_formats.py`](examples/output_formats.py) |
| **Batch Processing** | Process multiple queries efficiently | [`examples/batch_processing.py`](examples/batch_processing.py) |
| **Multi-Loop Research** | Iterative research with refinement | [`examples/multi_loop_research.py`](examples/multi_loop_research.py) |
| **Session Management** | Conversation tracking and persistence | [`examples/session_management.py`](examples/session_management.py) |
| **Chat Interface** | Interactive web-based chat demo | [`examples/chat_demo.py`](examples/chat_demo.py) |
**Quick Start Examples:**
```bash
# Basic research
python examples/basic_usage.py
# With export functionality
python examples/export_example.py
# Advanced configuration
python examples/advanced_config.py
```
**[View All Examples →](examples/README.md)**
## Key Features
| Feature | Description |
|---------|-------------|
| **Orchestrator-Worker Architecture** | A `Director Agent` coordinates research strategy while specialized worker agents execute focused search tasks with Exa API integration. |
| **Advanced Web Search Integration** | Utilizes `exa_search` with structured JSON responses, content summarization, and intelligent result extraction for comprehensive research. |
| **High-Performance Parallel Execution** | Leverages `ThreadPoolExecutor` to run multiple specialized agents concurrently, achieving significant time reduction for complex queries. |
| **Flexible Configuration** | Customizable model selection (Claude, GPT), token limits, loop counts, and output formatting options. |
| **Conversation Management** | Built-in conversation history tracking with the `swarms` framework's `Conversation` class for persistent dialogue management. |
| **Export Functionality** | JSON export with automatic timestamping, unique session IDs, and comprehensive conversation history. |
| **Multiple Output Formats** | Support for various output types including JSON, markdown, and full conversation history formatting. |
| **Session Management** | Unique session IDs, batch processing capabilities, and step-by-step research execution control. |
## Architecture
The system follows a streamlined orchestrator-worker pattern with parallel execution:
```
[User Query + Configuration]
│
▼
┌─────────────────────────────────┐
│ AdvancedResearch │ (Main Orchestrator)
│ - Session Management │
│ - Conversation History │
│ - Export Control │
└─────────────────────────────────┘
│ 1. Initialize Research Session
▼
┌─────────────────────────────────┐
│ Director Agent │ (Research Coordinator)
│ - Query Analysis & Planning │
│ - Task Decomposition │
│ - Research Strategy │
└─────────────────────────────────┘
│ 2. Decompose into Sub-Tasks
▼
┌─────────────────────────────────────────┐
│ Parallel Worker Execution │
│ (ThreadPoolExecutor - Concurrent) │
└─────────────────────────────────────────┘
│ │ │ │
▼ ▼ ▼ ▼
┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐
│Worker 1 │ │Worker 2 │ │Worker 3 │ │Worker N │
│Exa Search│ │Exa Search│ │Exa Search│ │Exa Search│
│Integration│ │Integration│ │Integration│ │Integration│
└──────────┘ └──────────┘ └──────────┘ └──────────┘
│ │ │ │
▼ ▼ ▼ ▼
┌─────────────────────────────────────────┐
│ Results Aggregation │
│ - Combine Worker Outputs │
│ - Format Research Findings │
└─────────────────────────────────────────┘
│ 3. Synthesize Results
▼
┌─────────────────────────────────┐
│ Conversation Management │
│ - History Tracking │
│ - Output Formatting │
│ - Export Processing │
└─────────────────────────────────┘
│ 4. Deliver Results
▼
[Formatted Report + Optional JSON Export]
```
### Workflow Process
1. **Session Initialization**: `AdvancedResearch` creates a unique research session with conversation tracking
2. **Director Agent Planning**: The director agent analyzes the query and plans research strategy
3. **Parallel Worker Execution**: Multiple worker agents execute concurrent searches using Exa API
4. **Results Aggregation**: Worker outputs are combined and synthesized into comprehensive findings
5. **Output Processing**: Results are formatted according to specified output type (JSON, markdown, etc.)
6. **Export & Delivery**: Optional JSON export with timestamped files and conversation history
## Contributing
This implementation is part of the open-source `swarms` ecosystem. We welcome contributions!
1. Fork the [repository](https://github.com/The-Swarm-Corporation/AdvancedResearch)
2. Create a feature branch (`git checkout -b feature/amazing-research-feature`)
3. Commit your changes (`git commit -m 'Add amazing feature'`)
4. Push to the branch (`git push origin feature/amazing-research-feature`)
5. Open a Pull Request
### Development Setup with uv
```bash
# Clone and setup development environment
git clone https://github.com/The-Swarm-Corporation/AdvancedResearch.git
cd AdvancedResearch
uv venv
uv pip install -r requirements.txt
# Setup environment variables
cp .env.example .env
# Edit .env with your API keys
```
## 📄 License
This project is licensed under the MIT License. See the [LICENSE](https://github.com/The-Swarm-Corporation/AdvancedResearch/blob/main/LICENSE) file for details.
## 📚 Citation
If you use this work in your research, please cite both the original paper and this implementation:
```bibtex
@misc{anthropic2024researchsystem,
title={How we built our multi-agent research system},
author={Anthropic},
year={2024},
month={June},
url={https://www.anthropic.com/engineering/built-multi-agent-research-system}
}
@software{advancedresearch2024,
title={AdvancedResearch: Enhanced Multi-Agent Research System},
author={The Swarm Corporation},
year={2024},
url={https://github.com/The-Swarm-Corporation/AdvancedResearch},
note={Implementation based on Anthropic's multi-agent research system paper}
}
@software{swarms_framework,
title={Swarms: An Open-Source Multi-Agent Framework},
author={Kye Gomez},
year={2023},
url={https://github.com/kyegomez/swarms}
}
```
## Documentation
For comprehensive API documentation, examples, and advanced usage:
**[View Full API Documentation →](DOCS.md)**
## Related Work
- [Original Paper](https://www.anthropic.com/engineering/built-multi-agent-research-system) - "How we built our multi-agent research system" by Anthropic
- [Swarms Framework](https://github.com/kyegomez/swarms) - The underlying multi-agent AI orchestration framework
## Support
- **Issues**: [GitHub Issues](https://github.com/The-Swarm-Corporation/AdvancedResearch/issues)
- **Discussions**: [GitHub Discussions](https://github.com/The-Swarm-Corporation/AdvancedResearch/discussions)
- **Community**: [Join our Discord](https://discord.gg/EamjgSaEQf)
<p align="center">
<strong>Built with <a href="https://github.com/kyegomez/swarms">Swarms</a> framework for production-grade agentic applications </strong>
</p>
Raw data
{
"_id": null,
"home_page": "https://github.com/The-Swarm-Corporation/AdvancedResearch",
"name": "advanced_research",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.10",
"maintainer_email": null,
"keywords": "artificial intelligence, multi-agent systems, research automation, anthropic, orchestrator-worker pattern, parallel execution, web search, citation generation, llm-as-judge",
"author": "Kye Gomez",
"author_email": "kye@swarms.world",
"download_url": "https://files.pythonhosted.org/packages/a1/ee/f2394c133bbdc91485bbb653592066d2c3f7694e1bfd6a5ee115a302888a/advanced_research-0.2.0.tar.gz",
"platform": null,
"description": "\n\n# Advanced Research System (Based on Anthropic's Paper)\n\n[](https://discord.gg/EamjgSaEQf) [](https://www.youtube.com/@kyegomez3242) [](https://www.linkedin.com/in/kye-g-38759a207/) [](https://x.com/kyegomezb)\n\n[](https://badge.fury.io/py/advanced_research)\n[](https://www.python.org/downloads/release/python-3100/)\n[](https://opensource.org/licenses/MIT)\n\nAn enhanced implementation of the orchestrator-worker pattern from Anthropic's paper, [\"How we built our multi-agent research system\"](https://www.anthropic.com/engineering/built-multi-agent-research-system), built on top of the bleeding-edge multi-agent framework `swarms`. Our implementation of this advanced research system leverages parallel execution, LLM-as-judge evaluation, and professional report generation with export capabilities.\n\n\n\n## Installation\n\n```bash\npip3 install -U advanced-research\n\n# uv pip install -U advanced-research\n```\n\n## Environment Variables\n\n```bash\n# Exa Search API Key (Required for web search functionality)\nEXA_API_KEY=\"your_exa_api_key_here\"\n\n# Anthropic API Key (For Claude models)\nANTHROPIC_API_KEY=\"your_anthropic_api_key_here\"\n\n# OpenAI API Key (For GPT models) \nOPENAI_API_KEY=\"your_openai_api_key_here\"\n\n# Worker Agent Configuration\nWORKER_MODEL_NAME=\"gpt-4.1\"\nWORKER_MAX_TOKENS=8000\n\n# Exa Search Configuration\nEXA_SEARCH_NUM_RESULTS=2\nEXA_SEARCH_MAX_CHARACTERS=100\n```\n\n**Note:** At minimum, you need `EXA_API_KEY` for web search functionality. For LLM functionality, you need either `ANTHROPIC_API_KEY` or `OPENAI_API_KEY`.\n\n## Quick Start\n\n### Basic Usage\n\n```python\nfrom advanced_research import AdvancedResearch\n\n# Initialize the research system\nresearch_system = AdvancedResearch(\n name=\"AI Research Team\",\n description=\"Specialized AI research system\",\n max_loops=1,\n)\n\n# Run research and get results\nresult = research_system.run(\n \"What are the latest developments in quantum computing?\"\n)\nprint(result)\n```\n\n### With Export Functionality\n\n```python\nfrom advanced_research import AdvancedResearch\n\n# Initialize with export enabled\nresearch_system = AdvancedResearch(\n name=\"Quantum Computing Research\",\n description=\"Research team focused on quantum computing advances\",\n max_loops=1,\n export_on=True, # Enable JSON export\n)\n\n# Run research - will automatically export to JSON file\nresearch_system.run(\n \"What are the latest developments in quantum computing?\"\n)\n# Results will be saved to a timestamped JSON file\n```\n\n### Advanced Configuration\n\n```python\nfrom advanced_research import AdvancedResearch\n\n# Initialize with custom settings\nresearch_system = AdvancedResearch(\n name=\"Medical Research Team\",\n description=\"Specialized medical research system\",\n director_model_name=\"claude-3-5-sonnet-20250115\", # Use latest Claude model\n worker_model_name=\"claude-3-5-sonnet-20250115\",\n director_max_tokens=10000,\n max_loops=2, # Multiple research iterations\n output_type=\"all\", # Include full conversation history\n export_on=True,\n)\n\n# Run research with image input (if applicable)\nresult = research_system.run(\n \"What are the most effective treatments for Type 2 diabetes?\",\n img=None # Optional image input\n)\n```\n\n### Batch Processing Multiple Queries\n\n```python\nfrom advanced_research import AdvancedResearch\n\n# Initialize the system\nresearch_system = AdvancedResearch(\n name=\"Batch Research System\",\n max_loops=1,\n export_on=True,\n)\n\n# Process multiple research tasks\ntasks = [\n \"Latest advances in renewable energy storage\",\n \"Current state of autonomous vehicle technology\",\n \"Recent breakthroughs in cancer immunotherapy\"\n]\n\n# Run batch processing\nresearch_system.batched_run(tasks)\n```\n\n### Using Different Output Formats\n\n```python\nfrom advanced_research import AdvancedResearch\n\n# Initialize with specific output type\nresearch_system = AdvancedResearch(\n name=\"Research System\",\n output_type=\"json\", # Options: \"all\", \"json\", \"markdown\"\n export_on=False, # Get results directly instead of exporting\n)\n\n# Run research and get formatted output\nresult = research_system.run(\n \"What are the key challenges in AGI development?\"\n)\n\n# Check available output methods\navailable_formats = research_system.get_output_methods()\nprint(f\"Available output formats: {available_formats}\")\n```\n\n## Quick Reference\n\n| Task | Code | Documentation |\n|------|------|---------------|\n| **Basic Research** | `AdvancedResearch().run(\"query\")` | [Basic Usage \u2192](DOCS.md#basic-research-setup) |\n| **Export Results** | `AdvancedResearch(export_on=True)` | [Export Config \u2192](DOCS.md#configuration-examples) |\n| **Batch Processing** | `system.batched_run([queries])` | [Batch Processing \u2192](DOCS.md#batch-processing-setup) |\n| **Custom Models** | `AdvancedResearch(director_model_name=\"model\")` | [Advanced Config \u2192](DOCS.md#advanced-multi-loop-research) |\n| **Output Formats** | `AdvancedResearch(output_type=\"json\")` | [Output Types \u2192](DOCS.md#types-and-enums) |\n\n## Examples\n\nReady-to-run examples demonstrating all features of the Advanced Research system:\n\n| Example | Description | File |\n|---------|-------------|------|\n| **Basic Usage** | Simple research with minimal configuration | [`examples/basic_usage.py`](examples/basic_usage.py) |\n| **Export Functionality** | Save research results to JSON files | [`examples/export_example.py`](examples/export_example.py) |\n| **Advanced Configuration** | Custom models, tokens, and multiple loops | [`examples/advanced_config.py`](examples/advanced_config.py) |\n| **Custom Models** | Different AI model configurations | [`examples/custom_models.py`](examples/custom_models.py) |\n| **Output Formats** | JSON, markdown, and conversation history | [`examples/output_formats.py`](examples/output_formats.py) |\n| **Batch Processing** | Process multiple queries efficiently | [`examples/batch_processing.py`](examples/batch_processing.py) |\n| **Multi-Loop Research** | Iterative research with refinement | [`examples/multi_loop_research.py`](examples/multi_loop_research.py) |\n| **Session Management** | Conversation tracking and persistence | [`examples/session_management.py`](examples/session_management.py) |\n| **Chat Interface** | Interactive web-based chat demo | [`examples/chat_demo.py`](examples/chat_demo.py) |\n\n**Quick Start Examples:**\n```bash\n# Basic research\npython examples/basic_usage.py\n\n# With export functionality \npython examples/export_example.py\n\n# Advanced configuration\npython examples/advanced_config.py\n```\n\n**[View All Examples \u2192](examples/README.md)**\n\n## Key Features\n\n| Feature | Description |\n|---------|-------------|\n| **Orchestrator-Worker Architecture** | A `Director Agent` coordinates research strategy while specialized worker agents execute focused search tasks with Exa API integration. |\n| **Advanced Web Search Integration** | Utilizes `exa_search` with structured JSON responses, content summarization, and intelligent result extraction for comprehensive research. |\n| **High-Performance Parallel Execution** | Leverages `ThreadPoolExecutor` to run multiple specialized agents concurrently, achieving significant time reduction for complex queries. |\n| **Flexible Configuration** | Customizable model selection (Claude, GPT), token limits, loop counts, and output formatting options. |\n| **Conversation Management** | Built-in conversation history tracking with the `swarms` framework's `Conversation` class for persistent dialogue management. |\n| **Export Functionality** | JSON export with automatic timestamping, unique session IDs, and comprehensive conversation history. |\n| **Multiple Output Formats** | Support for various output types including JSON, markdown, and full conversation history formatting. |\n| **Session Management** | Unique session IDs, batch processing capabilities, and step-by-step research execution control. |\n\n## Architecture\n\nThe system follows a streamlined orchestrator-worker pattern with parallel execution:\n\n```\n [User Query + Configuration]\n \u2502\n \u25bc\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502 AdvancedResearch \u2502 (Main Orchestrator)\n \u2502 - Session Management \u2502\n \u2502 - Conversation History \u2502\n \u2502 - Export Control \u2502\n \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n \u2502 1. Initialize Research Session\n \u25bc\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502 Director Agent \u2502 (Research Coordinator)\n \u2502 - Query Analysis & Planning \u2502\n \u2502 - Task Decomposition \u2502\n \u2502 - Research Strategy \u2502\n \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n \u2502 2. Decompose into Sub-Tasks\n \u25bc\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502 Parallel Worker Execution \u2502\n \u2502 (ThreadPoolExecutor - Concurrent) \u2502\n \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n \u2502 \u2502 \u2502 \u2502\n \u25bc \u25bc \u25bc \u25bc\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502Worker 1 \u2502 \u2502Worker 2 \u2502 \u2502Worker 3 \u2502 \u2502Worker N \u2502\n \u2502Exa Search\u2502 \u2502Exa Search\u2502 \u2502Exa Search\u2502 \u2502Exa Search\u2502\n \u2502Integration\u2502 \u2502Integration\u2502 \u2502Integration\u2502 \u2502Integration\u2502\n \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n \u2502 \u2502 \u2502 \u2502\n \u25bc \u25bc \u25bc \u25bc\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502 Results Aggregation \u2502\n \u2502 - Combine Worker Outputs \u2502\n \u2502 - Format Research Findings \u2502\n \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n \u2502 3. Synthesize Results\n \u25bc\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502 Conversation Management \u2502\n \u2502 - History Tracking \u2502\n \u2502 - Output Formatting \u2502\n \u2502 - Export Processing \u2502\n \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n \u2502 4. Deliver Results\n \u25bc\n [Formatted Report + Optional JSON Export]\n```\n\n### Workflow Process\n\n1. **Session Initialization**: `AdvancedResearch` creates a unique research session with conversation tracking\n2. **Director Agent Planning**: The director agent analyzes the query and plans research strategy\n3. **Parallel Worker Execution**: Multiple worker agents execute concurrent searches using Exa API\n4. **Results Aggregation**: Worker outputs are combined and synthesized into comprehensive findings\n5. **Output Processing**: Results are formatted according to specified output type (JSON, markdown, etc.)\n6. **Export & Delivery**: Optional JSON export with timestamped files and conversation history\n\n\n## Contributing\n\nThis implementation is part of the open-source `swarms` ecosystem. We welcome contributions!\n\n1. Fork the [repository](https://github.com/The-Swarm-Corporation/AdvancedResearch)\n2. Create a feature branch (`git checkout -b feature/amazing-research-feature`)\n3. Commit your changes (`git commit -m 'Add amazing feature'`)\n4. Push to the branch (`git push origin feature/amazing-research-feature`)\n5. Open a Pull Request\n\n### Development Setup with uv\n\n```bash\n# Clone and setup development environment\ngit clone https://github.com/The-Swarm-Corporation/AdvancedResearch.git\ncd AdvancedResearch\n\nuv venv\n\nuv pip install -r requirements.txt\n\n# Setup environment variables\ncp .env.example .env\n# Edit .env with your API keys\n```\n\n## \ud83d\udcc4 License\n\nThis project is licensed under the MIT License. See the [LICENSE](https://github.com/The-Swarm-Corporation/AdvancedResearch/blob/main/LICENSE) file for details.\n\n## \ud83d\udcda Citation\n\nIf you use this work in your research, please cite both the original paper and this implementation:\n\n```bibtex\n@misc{anthropic2024researchsystem,\n title={How we built our multi-agent research system},\n author={Anthropic},\n year={2024},\n month={June},\n url={https://www.anthropic.com/engineering/built-multi-agent-research-system}\n}\n\n@software{advancedresearch2024,\n title={AdvancedResearch: Enhanced Multi-Agent Research System},\n author={The Swarm Corporation},\n year={2024},\n url={https://github.com/The-Swarm-Corporation/AdvancedResearch},\n note={Implementation based on Anthropic's multi-agent research system paper}\n}\n\n@software{swarms_framework,\n title={Swarms: An Open-Source Multi-Agent Framework},\n author={Kye Gomez},\n year={2023},\n url={https://github.com/kyegomez/swarms}\n}\n```\n\n## Documentation\n\nFor comprehensive API documentation, examples, and advanced usage:\n\n**[View Full API Documentation \u2192](DOCS.md)**\n\n## Related Work\n\n- [Original Paper](https://www.anthropic.com/engineering/built-multi-agent-research-system) - \"How we built our multi-agent research system\" by Anthropic\n- [Swarms Framework](https://github.com/kyegomez/swarms) - The underlying multi-agent AI orchestration framework\n\n## Support\n\n- **Issues**: [GitHub Issues](https://github.com/The-Swarm-Corporation/AdvancedResearch/issues)\n- **Discussions**: [GitHub Discussions](https://github.com/The-Swarm-Corporation/AdvancedResearch/discussions)\n- **Community**: [Join our Discord](https://discord.gg/EamjgSaEQf)\n\n<p align=\"center\">\n <strong>Built with <a href=\"https://github.com/kyegomez/swarms\">Swarms</a> framework for production-grade agentic applications </strong>\n</p>\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "Advanced Multi-Agent Research System - Enhanced implementation of Anthropic's orchestrator-worker pattern with 90.2% performance improvement over single-agent systems",
"version": "0.2.0",
"project_urls": {
"Documentation": "https://github.com/The-Swarm-Corporation/AdvancedResearch/blob/main/Docs.md",
"Homepage": "https://github.com/The-Swarm-Corporation/AdvancedResearch",
"Repository": "https://github.com/The-Swarm-Corporation/AdvancedResearch"
},
"split_keywords": [
"artificial intelligence",
" multi-agent systems",
" research automation",
" anthropic",
" orchestrator-worker pattern",
" parallel execution",
" web search",
" citation generation",
" llm-as-judge"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "c5587d4e681cfc8c76e049f8d71598d17de56b6207a27542770395d56ded5efb",
"md5": "c1a0f060b592d027d5b0214977cc03a4",
"sha256": "312d2115243bd52071abaec9198ad8b53ea615f71869920376832c5a43cd04c3"
},
"downloads": -1,
"filename": "advanced_research-0.2.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "c1a0f060b592d027d5b0214977cc03a4",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.10",
"size": 16339,
"upload_time": "2025-08-23T17:16:06",
"upload_time_iso_8601": "2025-08-23T17:16:06.773789Z",
"url": "https://files.pythonhosted.org/packages/c5/58/7d4e681cfc8c76e049f8d71598d17de56b6207a27542770395d56ded5efb/advanced_research-0.2.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "a1eef2394c133bbdc91485bbb653592066d2c3f7694e1bfd6a5ee115a302888a",
"md5": "68daed8affbb32392d28ec17650e7470",
"sha256": "aa5126c32289fa4e810d36b72a7b88e7a68f579a9daba46aa21a6d5e50b05e89"
},
"downloads": -1,
"filename": "advanced_research-0.2.0.tar.gz",
"has_sig": false,
"md5_digest": "68daed8affbb32392d28ec17650e7470",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.10",
"size": 18810,
"upload_time": "2025-08-23T17:16:07",
"upload_time_iso_8601": "2025-08-23T17:16:07.696104Z",
"url": "https://files.pythonhosted.org/packages/a1/ee/f2394c133bbdc91485bbb653592066d2c3f7694e1bfd6a5ee115a302888a/advanced_research-0.2.0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-08-23 17:16:07",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "The-Swarm-Corporation",
"github_project": "AdvancedResearch",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"requirements": [
{
"name": "swarms",
"specs": []
},
{
"name": "loguru",
"specs": []
},
{
"name": "pydantic",
"specs": []
},
{
"name": "httpx",
"specs": []
},
{
"name": "python-dotenv",
"specs": []
},
{
"name": "requests",
"specs": []
},
{
"name": "orjson",
"specs": []
},
{
"name": "uvicorn",
"specs": []
},
{
"name": "black",
"specs": []
},
{
"name": "ruff",
"specs": []
},
{
"name": "swarms-tools",
"specs": []
}
],
"lcname": "advanced_research"
}
JSON
