What problem does it solve? Online research often produces conclusions built on secondhand summaries, outdated claims, and unverified numbers. This Skill enforces a disciplined research methodology so every factual statement traces back to a primary source and every claim carries an honest confidence level. ## Core Features & Use Cases - Primary-source tracing: Ranks sources by reliability (official docs, source code, papers first) and follows citation chains back to the original statement instead of trusting retellings. - Negative-claim caution: Treats "not found" differently from "does not exist", requiring bilingual and multi-phrasing searches before any negative assertion. - Cross-validation and confidence tiers: Requires at least two independent sources for key conclusions and separates confirmed facts, inferences, and guesses with distinct wording. - Use Case: When evaluating whether a framework supports a specific feature, the Skill checks the current official documentation and changelog rather than relying on a year-old blog post, and labels the conclusion with its source and confidence level. ## Quick Start Use the web-research skill to investigate the current state of plugin architectures in AI agent frameworks and cite primary sources for every claim.
