Curated, verified, analyzed.
No articles yet
Every article is sourced from a registry of 28 news outlets across three trust tiers. Tier 1 sources (Reuters, AP, BBC, NHK, ABC Australia) are prioritized for their editorial standards and fact-checking processes. We supplement RSS feeds with targeted searches for breaking developments, particularly in the Australia-Japan corridor.
We aim for balanced coverage across six categories: Technology, Business, Japan, Australia, World, and Digital. Stories are selected for newsworthiness, verifiability, and relevance.
Each article is evaluated for linguistic bias by examining framing, loaded language, sourcing balance, and headline-content alignment. We strip away the subject matter and assess the raw rhetorical structure.
Articles receive a bias direction (Left to Right) and a score from -3.0 to +3.0, where 0 represents neutral reporting. This helps readers identify potential editorial slant regardless of the topic.
For each article, we identify the 2-3 most significant factual claims and cross-reference them against independent sources. Articles are rated:
We scan each article for seven common logical fallacies: Appeal to Authority, Appeal to Majority, Slippery Slope, Whataboutism, Circular Reasoning, False Dichotomy, and Straw Man arguments. When detected, these are flagged so readers can evaluate the reasoning independently.
All external content is processed through a quarantined research pipeline. Web content is fetched by a sandboxed agent with no ability to take actions, write files, or access internal systems. The analysis agent never sees raw HTML. Content is sanitized, structured into validated JSON, and only then published. This architecture prevents prompt injection and ensures editorial integrity.