How was this work created?
Pick the path that best matches your process. You can always start over.
AI-assisted semantic analysis
On by default. When on, an LLM (Lovable AI, Gemini 3 Flash) reads your text and produces four semantic scores that blend into the rule-based engine. The tag is always chosen by transparent rules — never by the AI — and you'll see the AI's evidence quotes on the results page.
On — your text will be sent to Lovable AI for one analysis call, then discarded. Submissions longer than ~6,000 words per stage are summarized for the AI. Learn more
How the analysis works: a transparent, rule-based engine blends your text versions (when provided), your answers to the ten process questions, and any optional notes into four dimension scores that map to a taxonomy tag. When semantic analysis is on, an LLM contributes four additional origin/survival scores with evidence quotes — its output is shown to you and never chooses the tag.