Structural Entropy Check

Measure token bloat and find exactly which DOM regions are drowning out your content for AI agents.

Analyzing structural entropy…

— Entropy

Noise Ratio: — SNR: — Total Tokens: — Signal: — · Noise: —

🔥 Code-Noise Heatmap

📊 Token Distribution by DOM Region

RegionTokens% of TotalType

🛠 Automated Remediation Suggestions

📐 Methodology

Structural Entropy Check decomposes your page's HTML into DOM regions and measures the ratio of semantic signal to structural noise.

S = TITLE + META + H1 + CONTENT + JSONLD (semantic tokens) N = NAV + FOOTER + SCRIPTS + STYLES + ATTRS + HEADER + ASIDE (structural tokens) T = S + N NoiseRatio = N / T (0–1; lower is better) H = −Σ pᵢ log₂(pᵢ) (Shannon entropy over regions) NormalizedEntropy = H / log₂(k) (0–1; token distribution spread) EntropyScore = 100 × (1 − 0.6×(1−S/T) − 0.4×(1−NormalizedEntropy))

Thresholds: Good ≥ 80 · Fair 50–79 · Poor < 50

Token Bloat: Low (NoiseRatio < 0.4) · Medium (0.4–0.69) · High (≥ 0.7)

Full Methodology & Whitepaper →