AI literacy consensus tool

About this app

We created a live tool for this participatory workshop to support the analysis of participants’ responses, given the large room and the need for quick thematic analysis. This involved vibe coding.

The figure below shows how the tool was built, how each stage of the workshop runs, and how a scoping review informs the clustering.

Figure in three panels. A: the tool was built by the research team with Claude Sonnet 5.5 through Claude Code, stored on GitHub and deployed by Vercel. B: the five workshop stages (presentation, ideation, clustering and review, prioritisation, results) with the prompts and parameters used at each. C: the scoping review's 17 competency statements and the participants' responses both go into a Claude clustering prompt, with the review used only as a lens, producing thematic clusters for the facilitators to edit into the ballot.

Download the figure: PNG · SVG