Reading is not comprehension.
Comprehension is not capability.
Deepread closes all three gaps.
A five-layer annotation protocol that transforms reading into demonstrated capability. Every text enters as words and exits as feats.
14Texts in Corpus
13Fully Annotated
63Feats Generated
10Ladders Designed
Text Mode
Single Book
Full L1-L5 annotation of one text. The question, the summary, the x-ray, the decomposition, the premises. Produces feats.
Course Mode
Domain Mastery
Multiple texts in sequence, building toward a ladder. Cross-read synthesis, productive tensions, domain frameworks.
Scan Mode
Orientation
30-minute structural scan. TOC x-ray, core claim extraction, read/skim/skip verdict. Triage before commitment.

The Protocol

L1
The Question
Not what the book covers. The animating question the author could not stop thinking about.
L2
Authorial Summary
A summary the author would sign. Note one thing you were tempted to project that you removed.
L3
Structural X-Ray
Table of sections: CORE, SUPPORT, SCAFFOLDING. Where does the argument actually live? What could be cut?
L4
Claim Decomposition
[C] claim, [E] evidence, [ex] example, [R] rhetorical. Where does rhetoric substitute for evidence?
L5
Premise Excavation
Three premises the author relies on but does not defend. For each: what fails without it?