Sharp tools for sharp thinkers

Scans your article paragraph-by-paragraph for weak endings, dead weight, and energy dips. Pure NLP -- no AI API needed.

Measures concreteness, emotional intensity, and semantic tension using psycholinguistic datasets. No AI -- pure NLP.

Creates title candidates scored by engagement patterns learned from 1,200+ analyzed articles.

7-stage math-first pipeline: SWOW cliche avoidance, GloVe surprise gaps, Warriner emotion scoring, bisociation mixing. The LLM just formats what the math found.

Uses GloVe embeddings to find high-tension word collisions across semantic domains. Arthur Koestler's bisociation theory, computed.

Cross-joke emotional arc analysis, callback detection, island flagging, energy pacing, and algorithmic reordering. Pure NLP.

BFS through knowledge graphs + GloVe embeddings. 6-component scoring: surprise, inevitability, emotional charge, vividness, tonal shift. Zero LLM -- 100% math.

Challenges your hypothesis with counter-evidence, alternative explanations, and Bayesian reasoning. Uses Claude + Perplexity.

7 parallel attack angles (assumption, inversion, pre-mortem, reframe, deep research) synthesized into BUILD/RETHINK/KILL verdict. Not one prompt -- seven specialized ones + Perplexity grounding.

Two-tier pipeline: local scoring on 7 dimensions, then 3 separate deep research queries per idea with web-grounded data. Your full profile embedded.