• The Pointer: Auto-extract measurable skills from any syllabus and map them to a competency framework.
• The Details: Faculty write objectives in natural language; this pipeline converts them into structured, assessable competency units aligned to Bloom's or O*NET.
• Action:

gemini "Extract all learning objectives from this syllabus. Map each to a Bloom's taxonomy level and suggest one authentic assessment per objective. Output as JSON."

Phase 5: Institutional Leadership & AI Governance

Backlink: Dailies MOC

• The Pointer: Map connections between courses, departments, and skills across the institution using an AI-generated knowledge graph.
• The Details: When you operate at scale, context collapse kills quality. A graph lets the AI 'see' cross-departmental dependencies and prevent curriculum redundancy or gaps.
• Action:

gemini "Given these syllabi, generate a Mermaid graph showing shared competencies and prerequisite chains across CUIN, HIED, and SPSY programs."

Phase 5: Institutional Leadership & AI Governance

Backlink: Dailies MOC

• The Pointer: Quantify the accessibility gap in a course or program before proposing remediation.
• The Details: Use a structured rubric (WCAG 2.2, Section 508, UDL) to estimate the 'debt' - hours, cost, risk - of every non-conformant element. Makes the business case for upfront accessibility.
• Action:

gemini "Analyze this course shell. Output a JSON object with: { wcag_failures: [], estimate_hours: N, risk_level: 'high|medium|low' }"

Phase 5: Institutional Leadership & AI Governance

Backlink: Dailies MOC

• The Pointer: Treat AI usage policies not as static text, but as 'Guardrail Systems' injected into every prompt.
• The Details: As a Senior ID, you define the 'boundary' for faculty. Convert the UT AI Policy into a system prompt that can be applied to any course design task.
• Action:

gemini --system ./04_Policy/UT_AI_Guardrails.md "Review this assignment. Does it provide enough human-in-the-loop requirements for students?"

Phase 5: Institutional Leadership & AI Governance

Backlink: Dailies MOC

• The Pointer: End your 40-day cycle by asking the AI to audit your entire workspace_technical_manual.md against your actual 40-day log.
• The Details: Your system should learn from your habits. If you've been using a specific command often, it belongs in the official Manual.
• ID Application: Ensures your "Agentic Infrastructure" grows with you, preventing "Instructional Debt."
• Action: gemini "Based on my last 10 days of CLI Logs, suggest 3 updates to my Workspace Technical Manual to better reflect my current workflow."