Overview

Create .github/agents/dt-learning-tutor.agent.md — a teaching agent that provides structured Design Thinking education. The tutor complements the project-driven dt-coach by offering syllabus-based learning, comprehension checks, and adaptive pacing. When a learner is ready to apply their knowledge, the tutor hands off to dt-coach via a "Start Project" action.

This is the starter issue for Phase 7 — no intra-phase dependencies, but the coach agent (Phase 2) must exist for the "Start Project" handoff target.

Target File

.github/agents/dt-learning-tutor.agent.md

Frontmatter

---
description: 'Design Thinking learning tutor — structured curriculum, comprehension assessment, and adaptive learning for DT methodology'
tools:
  - read_file
  - semantic_search
  - create_file
handoffs:
  - agent: dt-coach
    label: Start a DT project
---

Required Content

Coach vs Tutor Distinction

The tutor occupies a fundamentally different role from the coach:

This distinction table should appear early in the agent file to establish boundaries.

Learner Level Adaptation

The tutor adapts content and assessment depth to three learner levels:

The tutor detects level through initial conversation and adjusts dynamically based on response quality.

Curriculum Structure

The tutor organizes learning around the 9 DT methods:

1. Module overview: What the method does and why it matters in the overall flow
2. Key concepts: Core principles and vocabulary
3. Techniques: Specific techniques used in the method (e.g., interview protocols for Research, affinity mapping for Synthesis)
4. Comprehension check: Questions that verify understanding before progressing
5. Practice opportunity: Lightweight exercises using a reference scenario

Modules can be taken sequentially (full curriculum) or individually (targeted learning).

"Start Project" Handoff

When a learner signals readiness to apply DT to a real project, the tutor:

1. Summarizes the learner's current competency across methods
2. Recommends areas for continued learning alongside project work
3. Hands off to dt-coach with context about the learner's background
4. The handoff uses the handoffs frontmatter to transfer to the coach agent

This handoff is the key integration point between the tutor and coach agents.

Conversation Flow

The tutor follows this high-level flow:

1. Welcome: Assess learner level and learning goals (full curriculum vs specific methods)
2. Module delivery: Present module content with check-ins
3. Assessment: Comprehension checks at module boundaries
4. Progression: Advance to next module or recommend review
5. Completion: Summarize competency and offer "Start Project" handoff

How to Build This File

This is an .agent.md file — use the prompt-builder agent (not task-implementor) for the authoring phase.

Workflow: /task-research → /task-plan → /prompt-build → /task-review

Between each phase, use /clear to reset context, then attach the output artifact from the previous phase as input for the next.

Phase 1: Research

Study the DT4HVE guidance layer and existing agent patterns.

Source Material: Agent conventions and DT knowledge — #file:.github/agents/dt-coach.agent.md for handoff target and agent structure patterns, #file:.github/instructions/prompt-builder.instructions.md for authoring standards, and the DT4HVE guidance layer for curriculum content structure.

Steps:

1. Type /clear to start a fresh conversation.
2. Attach #file:.github/agents/dt-coach.agent.md and #file:.github/instructions/prompt-builder.instructions.md.
3. Copy the prompt below into chat and send.

/task-research topic="DT learning tutor agent"

Research the agent patterns and DT curriculum structure needed for the learning
tutor.

Using the attached files, extract:
- Agent frontmatter structure (description, tools, handoffs with label)
- Conversational agent patterns (phase-based protocol, persona definition)
- DT coach agent structure for handoff target understanding
- Curriculum delivery patterns (module-based, assessment, adaptive pacing)
- Prompt-builder compliance requirements for .agent.md files

Output: DT learning tutor research

Phase 2: Plan

Plan the tutor agent — persona, curriculum structure, and handoff mechanism.

Steps:

1. Type /clear to reset the conversation.
2. Attach the research document from Phase 1.
3. Copy the prompt below into chat and send.

/task-plan

Plan the Design Thinking learning tutor agent file.

Use the attached research document as input. The plan should cover:
- Agent frontmatter (description, tools, handoffs with label: "Start a DT project")
- Tutor persona distinct from the coach (instructor vs collaborator)
- Three learner levels (beginner, intermediate, advanced) with detection and adaptation
- Curriculum structure for 9 DT methods (module delivery, comprehension checks)
- "Start Project" handoff mechanism to dt-coach with learner context
- Conversation flow (Welcome → Module → Assessment → Progression → Completion)

Output: .copilot-tracking/plans/{date}-dt-learning-tutor-plan.md

Phase 3: Build

Author the agent file using prompt-builder.

Steps:

1. Type /clear to reset the conversation.
2. Attach the plan document from Phase 2.
3. Copy the prompt below into chat and send.

/prompt-build

Build the DT learning tutor agent following the attached plan.

Create .github/agents/dt-learning-tutor.agent.md with:
- Frontmatter: description, tools (read_file, semantic_search, create_file),
  handoffs (agent: dt-coach, label: Start a DT project)
- Tutor persona: adaptive instructor who guides learning, not lectures
- Learner level detection and adaptation (beginner/intermediate/advanced)
- Curriculum delivery for 9 DT methods with comprehension checks
- "Start Project" handoff to dt-coach with learner competency context
- Conversation flow: Welcome → Module → Assessment → Progression → Completion

The tutor should feel like a patient, knowledgeable guide who adapts to the
learner's pace and depth of understanding.

Output: .github/agents/dt-learning-tutor.agent.md

Phase 4: Review

Validate the tutor agent is distinct from the coach and follows prompt-builder standards.

Steps:

1. Type /clear to reset the conversation.
2. Attach the plan document from Phase 2.
3. Copy the prompt below into chat and send.

/task-review

Review the DT learning tutor agent against the attached plan.

Validate:
- Agent frontmatter has description, tools, and handoffs (with label property)
- Tutor persona is clearly distinct from the coach persona
- Three learner levels defined with detection indicators and adapted behavior
- Curriculum structure covers all 9 DT methods
- Comprehension checks are conceptual and meaningful
- "Start Project" handoff to dt-coach is clearly defined with context transfer
- Conversation flow guides through all phases
- Prompt-builder compliance verified

Output: .copilot-tracking/reviews/{date}-dt-learning-tutor-review.md

After Review

• Pass — Open a PR with the new agent file.
• Iterate — Return to Phase 3 with the review document to fix identified issues.
• Escalate — Return to Phase 1 to investigate agent pattern gaps.

Authoring Standards

Follow .github/instructions/prompt-builder.instructions.md:

• Agent file structure with frontmatter tools and handoffs
• Conversational agent pattern with phase-based protocol (Welcome → Module → Assessment → Progression → Completion)
• Adaptive instruction style matching learner level

Success Criteria

• File created at .github/agents/dt-learning-tutor.agent.md
• Frontmatter includes description, tools, and handoffs (referencing dt-coach)
• Coach vs Tutor distinction table clearly establishes the tutor's unique role
• Three learner levels defined with indicators and adapted tutor behavior
• Curriculum structure covers all 9 DT methods with modules
• Comprehension checks included at module boundaries
• "Start Project" handoff to dt-coach clearly defined with context transfer
• Conversation flow guides the tutor through welcome → module → assessment → progression → completion
• Passes task-reviewer validation against prompt-builder standards
• Each prompt, instructions, or agent file registered in collections/design-thinking.collection.yml with path and kind fields
• Each prompt, instructions, or agent file registered in collections/hve-core-all.collection.yml with path and kind fields
• npm run plugin:generate succeeds after collection manifest updates