Make the rigorous path the easy path.

A controlled path from question to reviewable output.

Evidence Engine

A specification-first workflow that moves research through explicit planning, execution, and review gates. The framework is designed to make the rigorous path easier to follow and easier to audit.

• Define the intended artifact before drafting
• Keep analysis and source claims traceable
• Use reporting guidelines as working checklists

Magnitude and precision before threshold labels.

Estimation First

Point estimates, absolute effects, and 95% confidence intervals carry the interpretation. P-values, when present, remain supplementary rather than load-bearing.

• Report the effect in clinically readable units
• Interpret interval location and width
• Separate statistical compatibility from clinical importance

Guidelines scored and citations checked.

Verification Layer

Study-specific validators check STARD, STROBE, PRISMA-ScR, TRIPOD, and related requirements. References must resolve through authoritative sources; unverifiable claims are flagged rather than completed by guesswork.

• Use the correct reporting guideline
• Verify identifiers and quoted endpoints
• Surface limitations instead of softening them

Prediction, outcome, debrief, update.

Decision Journal

A private proof of concept for recording preoperative predictions, intraoperative pivots, resolved outcomes, and calibration over time. The public site demonstrates the idea without collecting clinical case data.

• Record predictions before outcomes are known
• Separate decision quality from outcome quality
• Use repeated feedback to recalibrate judgment

Honest probability in forms patients can use.

Patient Compass

GRADE-rated fact boxes and natural-frequency displays make baseline risk, uncertainty, and pathway dependence visible without promising an individualized prediction.

• Prefer natural frequencies to opaque percentages
• Show uncertainty and evidence certainty
• Use patient-facing artifacts to support—not replace—conversation