Anti-hallucination research skill for Claude Code — admits uncertainty, extracts direct quotes before analysis, cites every claim, retracts unverifiable statements. Based on Anthropic's official guardrail techniques. By TheGEOLab.net
Mandatory web-first research workflow for any agentic runtime. Three-tier grounding: official docs → research papers → community perspectives. Evidence-verification pipeline with evidence-prediction (not confidence tags). Integrates with findout CLI and skill4eyes. No answer from parametric knowledge alone.
Grounded Agentic Research demonstration: a one-day, verification-gated sweep of memory retention through insect metamorphosis — claim-level epistemic tiers, DOI/CrossRef-audited citations, refutation-first pass, and a synthesis paper deriving a novel falsifiable hypothesis. AI-directed, human-verified, honestly labeled.
Add a description, image, and links to the grounded-research topic page so that developers can more easily learn about it.
To associate your repository with the grounded-research topic, visit your repo's landing page and select "manage topics."