Projects and Creating Projects
The Projects feature in ChatFrame is designed to help you organize your work, especially when dealing with multiple sets of local files for Retrieval-Augmented Generation (RAG).
What are Projects?
A project acts as a container for:
- Indexed Files: A specific collection of your local files (PDFs, code, documents) that you want the AI to be able to reference.
- Vector Store: The local, dedicated vector index created from those files.
- Chat History: The conversation history associated with that specific project.
- Settings: Potentially, project-specific settings for LLM models or MCP server configurations.
Creating a New Project
Creating a project is the first step to leveraging ChatFrame's Local RAG capabilities for a specific task or codebase.
- Navigate to Projects: In the ChatFrame sidebar or settings, find the Projects section.
- Start New Project: Click the "Create New Project" or "+" button.
- Name the Project: Give your project a descriptive name (e.g., "Q3 Financials," "Backend API Refactor").
- Add Files: Select the local files and folders you wish to include in the project.
- Indexing: ChatFrame will automatically begin the local indexing process, converting your file content into a private vector store.
Benefits of Using Projects
- Organization: Keep different work streams separate (e.g., personal research vs. work code).
- Context Control: Ensure the AI only has access to the relevant set of files for a given conversation, improving the accuracy of RAG.
- Performance: Smaller, more focused projects can lead to faster vector search and retrieval times.