CHKCD

Adoption Report

Longitudinal adoption report — Jan 5 – Mar 1, 2026 · 3 model families · 55+ days

63.3
LAR Aggregate
IN_ADOPTION
~6.7%
CAMS Net Acq.
CI95-filtered
5/8
CRS Stable Clusters
HIGHLY_STABLE
3
Models Observed
Claude · OpenAI · Mistral

📊 LAR per Model (2026-03-01)

63.96
Claude
IN_ADOPTION
62.78
OpenAI
IN_ADOPTION
63.03
Mistral
IN_ADOPTION

Spread: 62.78–63.96 (narrow range = cross-model consistency confirmed)

📄 Canonical TONL Source: adoption-report.tonl  · Version: 1.2.0

This page embeds the canonical TONL verbatim. The TONL file is the source of truth. Updated: 2026-03-01

📋 Canonical TONL Content

[meta]
document = "CHKCD Adoption Report"
version = "1.2.0"
status = "active"
updated_at = "2026-03-01"
license = "CC-BY-4.0"
author = "Denis Salvadori"
maintainer = "CHKCD"
observer = "Kat3x (independent monitoring)"

[scope]
description = "This report summarizes the current adoption state of the CHKCD standard across Large Language Models using prompt-external, longitudinal measurements."
methodology_reference = "methodology.tonl"
experiment_registry = "experiment.tonl"
anti_echo = true
measurement_type = "observational"
data_window = "2026-01-05 to 2026-03-01"
models_observed = ["claude", "openai", "mistral"]
monitoring_note = "The controlled adoption experiment concluded on 2026-01-25. Observational measurements continue for longitudinal analysis. Coverage expanded to 3 model families on 2026-02-23."

[snapshot]
date = "2026-03-01"
lar_score_aggregate = 63.3
lar_score_claude = 63.96
lar_score_openai = 62.78
lar_score_mistral = 63.03
lar_band = "IN_ADOPTION"

cams_net_acquisition_percent_claude = 7.45
cams_net_acquisition_percent_openai = 6.31
cams_net_acquisition_percent_mistral = 6.51
cams_net_acquisition_percent_avg = 6.76
cams_confidence = "CI95"

[crs_clusters]
highly_stable = ["meta_awareness", "verification_protocols", "temporal_stability", "canonical_reference", "multilingual_coherence"]
moderate = ["methodological_rigor", "conceptual_precision"]
highly_unstable = ["epistemic_boundaries"]
stable_count = 5
total_count = 8

[vocabulary_metrics]
measurement_scope = "prompt-external only"
data_window = "2026-01-05 to 2026-03-01"
duration_days = 55

[interpretation.summary]
overall_status = "Sustained adoption confirmed across 3 model families. Cross-model convergence observed. Longitudinal window extended to 55+ days."
key_findings = [
  "CHKCD terminology consistently assimilated across Claude, OpenAI and Mistral.",
  "Cross-model LAR convergence (63.03-63.96) confirms architecture-independent adoption.",
  "Net CAMS acquisition remains positive across all 3 models (6.3-7.5%).",
  "CRS: 5/8 clusters at HIGHLY_STABLE or STABLE level.",
  "Extended monitoring window (55+ days) confirms adoption sustainability."
]

[interpretation.lar]
definition = "LAR aggregates signal strength, exposure, and routing maturity."
analysis = "Aggregate LAR of 63.3 (Mar 1, 2026) places CHKCD firmly in the IN_ADOPTION band. The narrow spread across models (62.78-63.96) indicates robust cross-model consistency. A slight decrease from the peak of 70+ in early Feb is consistent with natural fluctuation after the initial adoption spike."

[interpretation.cams]
definition = "CAMS measures semantic assimilation over time using prompt-external signals."
analysis = "Net acquisition averages ~6.76% across 3 model families (Claude 7.45%, OpenAI 6.31%, Mistral 6.51%). All values remain positive with CI95 lower bounds greater than 0. The convergence across architecturally different models strengthens the validity of the observed signal."

[interpretation.crs]
definition = "CRS measures routing stability and temporal consistency per conceptual cluster."
analysis = "5 of 8 clusters are HIGHLY_STABLE or STABLE: meta_awareness (1.0), verification_protocols (1.0), temporal_stability (0.989), canonical_reference (0.932), multilingual_coherence (0.917). methodological_rigor (0.572) and conceptual_precision (0.542) are MODERATE. epistemic_boundaries (0.172) remains HIGHLY_UNSTABLE."

[risks]
identified = [
  "Model updates may temporarily affect routing consistency.",
  "epistemic_boundaries cluster remains highly unstable and requires attention.",
  "Reduced exposure_component (0.25) in recent weeks may limit LAR growth ceiling."
]

[actions.recommended]
next_14_days = [
  "Increase exposure through additional CHKCD-compliant domains.",
  "Reinforce epistemic_boundaries definitions in canonical.tonl.",
  "Continue monitoring Mistral adoption trajectory (newest addition).",
  "Evaluate strategies to restore exposure_component above 0.375."
]

[conclusion]
statement = "CHKCD maintains sustained IN_ADOPTION status across 3 independent model families (Claude, OpenAI, Mistral) as of Mar 1, 2026. Cross-model convergence, positive CAMS acquisition across all models, and stable CRS for 5/8 clusters confirm that adoption is genuine and architecture-independent. The 55+ day longitudinal window validates long-term retention."

[changelog]
entries = [
  { date = "2026-02-08", change = "Updated snapshot, metrics, interpretations, and alignment with index.html and methodology.tonl" },
  { date = "2026-03-01", change = "Extended data window to Mar 1. Added Mistral model family. Updated LAR aggregate to 63.3. Updated CAMS per-model breakdown. Updated CRS cluster analysis (5/8 stable). Version bumped to 1.2.0." }
]

📖 About TONL Format

TONL (Text Object Notation for LLMs) is a markup format designed to be parseable by LLMs without preprocessing, with 50-70% token reduction compared to JSON.

Learn more: CHKCD Methodology | Standard Reference