## Overview

Create the foundational instruction files that define the DT coaching identity, quality constraints, method sequencing logic, and the first three method-tier instructions (Problem Space). These files form the ambient tier — always loaded when DT artifacts are open — plus the Method 1-3 tier instructions.

Deliverables

Key Design Decisions

• All ambient instructions use glob **/.copilot-tracking/dt/** for auto-loading.
• Method instructions use glob **/.copilot-tracking/dt/**/method-{NN}* for method-specific loading.
• The coaching identity embeds Think/Speak/Empower behaviorally — no explicit rule references.
• Quality constraints enforce lo-fi output in Methods 1-6 and progressive fidelity in Methods 7-9.
• Method sequencing supports non-linear iteration (Method 8 → Method 2 loops).

Authoring Standards

All instruction files follow the prompt-builder standards in .github/instructions/prompt-builder.instructions.md:

• Required description: frontmatter with applyTo: glob
• Writing style: guidance over commands, * bullets, **bold** for key concepts
• Progressive disclosure: metadata ~100 tokens → instructions <5,000 tokens
• Content nativized from DT4HVE source with manufacturing references retained (V1 approach)

Artifact creation uses the RPI pipeline: task-researcher gathers DT4HVE source material for each instruction file, task-planner sequences creation, task-implementor authors files, task-reviewer validates against coaching quality criteria and prompt-builder standards.

Dependencies

• No external dependencies — this is the foundation phase.
• Internal: All Method 1-3 instructions depend on the coaching identity file.

Success Criteria

• All 6 instruction files created with proper frontmatter and globs
• Token budgets verified within estimated ranges
• Think/Speak/Empower philosophy embedded behaviorally in coaching identity
• 4-level progressive hint engine specified
• Method instructions cover 3 sub-methods each (planning → execution → documentation)
• All files pass 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