vela

This capability leverages a local LLM to provide real-time writing suggestions and contextual prompts based on the user's input. It employs a retrieval-augmented generation (RAG) approach, allowing the model to pull relevant information from a local knowledge base, enhancing the creativity and coherence of the writing process. This local-first architecture ensures user privacy and data security, as all processing occurs on the user's machine without external data transmission.

This capability allows users to dynamically generate character profiles and plot outlines based on user-defined parameters. By integrating a local LLM with a structured input format, users can specify traits, motivations, and story arcs, which the model uses to create detailed character sketches and plot summaries. This structured approach helps maintain narrative consistency and depth.

This capability analyzes the user's writing style and provides feedback on elements such as tone, pacing, and readability, all while ensuring that the data remains local. By employing natural language processing techniques, it evaluates the text without sending any information to external servers, thus maintaining user confidentiality. The feedback is presented in an actionable format to help improve writing quality.

This capability allows users to conduct research by querying a local knowledge base for relevant information and integrating it seamlessly into their writing. The RAG architecture enables the model to fetch contextually relevant data, which can be incorporated into the narrative, enhancing the depth and authenticity of the writing. This integration is designed to be intuitive, allowing for smooth transitions between research and writing.

This capability provides users with customizable templates for different writing formats, such as chapters, scenes, or character sketches. Users can modify these templates to fit their specific needs, allowing for a more structured approach to novel writing. The templates are designed to be flexible, enabling users to adapt them as their writing evolves.