Daily Intelligence Brief
Five prioritized events, compact event awareness, two distinct implications for the highest-consequence incidents, a visible structured-technique audit and numbered source links.
Product evidence
Illustrative sample and benchmark method
These examples demonstrate structure and quality controls. They are not live intelligence and must not be treated as a current threat assessment.
Russian Narrative Brief
Separates the observed messaging act from the truth of its claims, then assesses framing, audience, convergence and likely narrative effects.
Chinese Narrative Brief
Tracks state-run narratives as messaging artifacts, highlights meaningful change and applies the same source-audited implication standard.
Benchmark method
Products are evaluated on event recall, factual-source coverage, unsupported-claim rate, factual/estimative separation, likelihood calibration, implication-chain completeness, source-link validity, contradiction handling and readability. Evaluation sets use archived source packs and fixed prompts. No comparative superiority claim is published until a blinded human review has enough samples and inter-rater agreement.
Review
What is checked
Publication response
Evidence
Every factual incident claim resolves to a trusted reference
Repair once or retain local product
Analysis
Two distinct, traceable implication chains with calibrated language
Withhold unsupported premium assessment
Cost
Provider-reported usage remains within $0.25 per released product
Stop or use local fallback