I’m a high school student (16yo) building this because I write fiction and hate managing "story wikis" on the side.
Most AI writing tools just generate text. I wanted something that acts like a compiler/linter for the narrative itself.
What it does: It runs parallel agents in the background while you write. It builds a graph of your story’s state (characters, inventory, antagonist locations, timeline) in real-time. If you write that a character has a scar in Chapter 1, and then say they have smooth skin in Chapter 10, the system flags it.
The Tech: We aren't just wrapping a standard LLM. We're using a low-latency model to handle multiple "reasoning" streams at once—one for plot, one for character consistency, one for tone. It feels instant (sub-200ms processing).
I’m looking for feedback on the "Story Map" visualization. Is it too cluttered?