DecipherIt is an AI-powered research assistant that leverages Bright Data's MCP Server infrastructure to provide unrestricted, intelligent web access for collecting and synthesizing information. The MCP component powers global web access, multi-source data ingestion, and AI-driven analysis across documents, URLs, and topics. Once the MCP server is running, you can query the system to generate AI-powered summaries, interact with your research material through Q&A, and export outputs like mindmaps and FAQs. The included backend (.mcp_server) orchestrates data collection, vector storage, and agent collaboration to deliver cohesive research notebooks.

Prerequisites:

• Python 3.12 or compatible runtime
• Git
• Access to a Bright Data MCP API key (or configured environment for MCP interaction)

1. Clone the repository: git clone https://github.com/mtwn105/decipher-research-agent.git cd decipher-research-agent

2. Create and activate a Python virtual environment: python -m venv venv source venv/bin/activate # macOS/Linux

   Windows: venv\Scripts\activate.bat

3. Install dependencies (adjust if a requirements.txt exists; otherwise install needed packages listed in setup): pip install -r requirements.txt

4. Configure environment variables (example): export BRIGHTDATA_API_KEY=your-brightdata-api-key export BASE_URL=https://your-deployment-base-url export DATABASE_URL=postgresql://user:pass@host:port/dbname export LOG_LEVEL=info

5. Run the MCP server (as defined in mcp_config): python -m decipher_research_agent.mcp_server

6. Verify the server is running and accessible via the MCP client tooling or API endpoints described in the project documentation.

Notes:

• Ensure network access to any external services required by the agent stack (vector DB, authentication providers, etc.).
• Adjust BASE_URL and DATABASE_URL to your deployment environment.

Tips and common issues:

• If the MCP server fails to initialize, check that your environment variables are loaded correctly and that the API keys are valid.
• For large research tasks, tune MAX_RETRIES, request timeouts, and vector DB connection pools in the configuration.
• Enable verbose logging (LOG_LEVEL=debug) temporarily when debugging agent behavior or data ingestion pipelines.
• If using Bright Data MCP, confirm you have the proper endpoint availability and license for web access capabilities.
• This project’s backend relies on Python 3.12 features; ensure your runtime matches.
• If migrating to a containerized environment, map the environment variables into the container and expose the appropriate ports for MCP communication.