![Author ORCID: We display the ORCID iD icon alongside authors names on our website to acknowledge that the ORCiD has been authenticated when entered by the user. To view the users ORCiD record click the icon. [opens in a new tab]](https://www.cambridge.org/engage/assets/public/coe/logo/orcid.png){kind=logo}
Abstract
Two 2026 contributions to biodesign-pedagogy literature treat AI at distinct layers of the curricular stack. Riffle and Rubery (2026) document "Biodesign Buddy," a domain-specific large language model used as a scientific-literacy tutor with 64 students; Ojeda-Ramirez, Hakim and Danies (2026), extending Walker and Kafai's (2021) constructionist tradition, treat student prompts to generative image tools as a "grammar of speculation." This paper offers a third position. Through analytic reading of three coordinated curricular artifacts developed for a single MDes cohort at the University of Macau preparing for the 2026 Biodesign Challenge (BDC), it argues that AI-integrated biodesign pedagogy is fruitfully analysed at the instructor-side rubric layer: the layer at which an instructor authors, deposits, and versions the artifacts that configure both the agentic AI tool's responses and the evaluative instrument under which student work is graded. The construct AI Process Literacy — articulated post-hoc as three sub-criteria (Provenance, Curation, Reasoning with prompts) graded on a four-point scale — is offered as one workable rubric-level operationalisation of that layer. The paper engages the journal's open Question, "Does biotech education need new teaching methodologies?" (Walker), and is anchored in Goodwin's professional vision and a refined reading of legitimate peripheral participation (Lave and Wenger 1991; Fuller and Unwin 2003). The construct is set alongside Perkins et al.'s (2024) AI Assessment Scale (permission layer) and Elkhodr and Gide's (2026) SAGE framework (process layer); the three are non-rivalrous. Evidence is reported descriptively from three cohort surveys (baseline N=29; end-of-course N=25; BDC-reflection N=4).
Supplementary materials
Title
Supplement 1: Eight-Step Ideation Workbook
Description
The eight-step ideation workbook documented in §4.1 of the manuscript: a structured guide with per-step "USE CURSOR" callouts naming the kind of agent move appropriate to each step, from interest-mapping and intersection statements through rubric stress-testing to a four-week sprint plan.
Actions
Title
Supplement 2: Agent Context File (biodesign_cursor_agent.md)
Description
The Cursor agent context file documented in §4.2 of the manuscript, loaded into the agent's runtime at the start of every student session: the agent's role delineation, the BDC judging rubric and 2.75 sufficiency threshold, past winners 2023–2025, the bio + digital four-quadrant map, and response guidelines. Deposited verbatim in its native markdown form.
Actions
Title
Supplement 3: Modified Course Rubric (Diff)
Description
The modified course rubric documented in §4.3 of the manuscript, presented as a diff against the BDC's standing rubric: language added to ten criteria across the four standing categories, plus the cross-cutting AI Process Literacy dimension (Provenance, Curation, Reasoning with prompts) with its 4/3/2/1 scoring anchors and the Disclosure precondition.
Actions
Title
Supplement 4: Student Reflection Excerpts (BDC-Reflection Survey, N=4)
Description
Anonymised excerpts from the BDC-reflection survey (N=4) cited in §5.2 of the manuscript, including the survey instrument, Likert responses, and the consent-confirmed open-ended responses (Students B–E). Administered 6–8 May 2026 with explicit student consent; approved by the University of Macau Research Ethics Committee (HE-1755-2026, 12 May 2026).
Actions
