Mochocyang/QMAI

What it solves

QMAI is a desktop writing system designed for long-form novels (2-3 million words), addressing the common failures of standard AI writing tools: memory loss of early plot points, inconsistent character personalities, chaotic timelines, and lost foreshadowing. It ensures the AI maintains a consistent world-state and narrative style across massive volumes of text.

How it works

The system uses a structured memory engine that automatically extracts and stores information from chapters after they are saved. This creates a local database of summaries, character states, relationship changes, and timeline events.

When generating new text, QMAI assembles a "context package" based on a priority hierarchy (e.g., current outline, previous chapter ending, character knowledge, and vector/graph search results). It also employs a "Soul System" to define the overall narrative tone and specific character "DNA" to ensure consistent voice and behavior.

Who it’s for

• Web novel authors who need to maintain quality and consistency in long-running series.
• Novel planners managing complex world-building, faction relationships, and multi-threaded plots.
• AI-assisted writers seeking a tool that remains functional over millions of words.
• Study writers who want to analyze and replicate the style and character archetypes of existing successful works.

Highlights

• Three-tier Outline System: Hierarchical organization from general plot to volume and chapter-level details with visualization.
• Hybrid Memory Retrieval: Combines recent windowing, keyword search (BM25), vector search, and graph-based relationship tracking.
• Character Cognition Tracking: Specifically tracks what each character knows vs. what they do not know to prevent "meta-knowledge" errors.
• De-AI Rewriting: Built-in tools to transform standardized AI prose into more natural, human-like writing styles.
• Work Analysis (拆文库): Ability to import existing novels to extract character skills and 9-dimensional style metrics for reuse.
• Multi-dimensional Review: An AI auditing system that scores chapters on "satisfaction density," pacing, and character consistency.
• Plot Simulation Room: A multi-agent environment to simulate character reactions to specific triggers and output reports.