Vibe Compiler: A Research-Logic Synthesis Tool That Runs without Prompt Engineering -Toward Enhancing Metacognition for Sustaining Agency in the Age of Generative AI-
This study addresses the risk that generative AI, when used to support research, may lead users to uncritically accept its outputs, thereby undermining human cognitive agency. To counter this, the authors propose a Synthesis-Analysis Reciprocity Model and implement it in the Vibe Compiler system, which translates researchers’ vague intuitions (“Vibes”) into rigorous logical structures via a 16-dimensional academic ontology. When logical gaps are detected, the system prompts users with reflective questions to autonomously complete the reasoning rather than relying on AI-generated content. The work innovatively categorizes four origins of logical discontinuities and introduces a mechanism whereby the AI actively interrogates its own outputs to stimulate human metacognition. Prototype evaluation demonstrates that effective AI-assisted reasoning depends more on structured knowledge inputs than on sophisticated prompt engineering, significantly reinforcing researchers’ critical agency in human-AI collaboration.