feat(instructions): add DT-to-RPI handoff contract specification (#679)

# Pull Request

## Description

Add ambient-tier instruction defining the formal contract for lateral
handoffs from Design Thinking coaching into the RPI (Research → Plan →
Implement) workflow.

The handoff contract includes:
- **Tiered Handoff Schema** — Three exit points aligned to DT space
boundaries (Problem Statement Complete, Concept Validated,
Implementation Spec Ready)
- **Exit-Point Artifact Schema** — YAML template for handoff summary
files extending the coaching state pattern
- **RPI Input Contracts** — Per-agent table defining what each RPI agent
(researcher, planner, implementor) needs from DT outputs
- **Graduation Awareness Behavior** — Four-step flow (Detect → Surface →
Prepare → Transfer) for the DT coach at space boundaries
- **Handoff Quality Markers** — Confidence taxonomy
(validated/assumed/unknown/conflicting) with RPI implications

Token budget: ~1,500–2,000 tokens (ambient tier, `applyTo:
'**/.copilot-tracking/dt/**'`).

## Related Issue(s)

Closes #583

## Type of Change

**AI Artifacts:**

- [x] Copilot instructions (`.github/instructions/*.instructions.md`)

## Sample Prompts (for AI Artifact Contributions)

**User Request:**
"We've finished our synthesis and have a validated problem statement.
Let's hand this off to the RPI workflow for implementation planning."

**Execution Flow:**
1. DT coach detects space boundary readiness signals (Method 3 complete,
synthesis validated).
2. Coach surfaces lateral handoff option: "Your problem statement looks
solid. You could continue to Method 4 (Brainstorming), revisit earlier
methods, or hand off to the RPI Researcher to investigate implementation
approaches."
3. If handoff chosen, coach creates `handoff-summary.md` with artifacts
tagged by confidence markers.
4. Coach records lateral transition in coaching state and provides
handoff summary path to the RPI agent.

**Output Artifacts:**
`.copilot-tracking/dt/{project-slug}/handoff-summary.md` —
YAML-structured handoff file with exit point, artifacts, constraints,
and assumptions, each tagged with confidence markers.

**Success Indicators:**
- Handoff summary contains all DT artifacts with confidence markers
- RPI agent receives structured input matching its input contract
- Coach remains available for advisory after handoff

## Testing

- `npm run lint:md` — 0 errors (183 files)
- `npm run lint:frontmatter` — 0 errors, 0 warnings
- `npm run lint:collections-metadata` — 11 collections validated, 0
errors
- `npm run plugin:generate` — 11 plugins generated successfully

## Checklist

### Required Checks

- [x] Documentation is updated (if applicable)
- [x] Files follow existing naming conventions
- [x] Changes are backwards compatible (if applicable)

### AI Artifact Contributions

- [x] Verified contribution follows common standards and type-specific
requirements

### Required Automated Checks

- [x] Markdown linting: `npm run lint:md`
- [x] Frontmatter validation: `npm run lint:frontmatter`

## Security Considerations

- [x] This PR does not contain any sensitive or NDA information

## Additional Notes

Part of Epic #563 — Design Thinking Coaching Capabilities (Phase 5:
Implementation Space). This ambient instruction enables the DT coach
agent to formally transition teams from coaching into the RPI
implementation workflow with structured artifact handoff.

---------

Co-authored-by: Bill Berry <wbery@microsoft.com>

Author: chaosdinosaur

.github/instructions/design-thinking/dt-rpi-handoff-contract.instructions.md

@@ -0,0 +1,90 @@
+---
+description: 'DT-to-RPI handoff contract defining exit points, artifact schemas, and per-agent input requirements for lateral transitions from Design Thinking to RPI workflow'
+applyTo: '**/.copilot-tracking/dt/**'
+---
+
+# DT→RPI Handoff Contract
+
+Defines the formal contract for lateral handoffs from Design Thinking coaching into the RPI (Research → Plan → Implement) workflow. Use this guidance whenever a team graduates from a DT space boundary or explicitly requests implementation support.
+
+## Tiered Handoff Schema
+
+Three exit points align with DT space boundaries. Each exit targets the RPI agent best suited to consume DT outputs at that stage.
+
+| Exit Point                 | DT Methods   | DT Space Boundary         | RPI Target  | What Transfers                                                                         |
+|----------------------------|--------------|---------------------------|-------------|----------------------------------------------------------------------------------------|
+| Problem Statement Complete | 1-3 complete | Problem → Solution        | Researcher  | Validated problem statement, synthesis themes, stakeholder map, constraint inventory   |
+| Concept Validated          | 4-6 complete | Solution → Implementation | Planner     | Tested concepts, lo-fi prototype feedback, constraint discoveries, narrowed directions |
+| Implementation Spec Ready  | 7-8 complete | Implementation exit       | Implementor | Hi-fi prototype specs, user testing results, architecture decisions, rollout criteria  |
+
+Earlier exit points produce more RPI work. A Problem Statement Complete handoff requires full RPI research and planning. An Implementation Spec Ready handoff may skip directly to implementation.
+
+## Exit-Point Artifact Schema
+
+Record handoff artifacts in the coaching state `transition_log` using a lateral transition entry. Create a handoff summary file alongside the coaching state.
+
+```yaml
+# .copilot-tracking/dt/{project-slug}/handoff-summary.md
+exit_point: "problem-statement-complete | concept-validated | implementation-spec-ready"
+dt_method: 3          # last completed DT method
+dt_space: "problem"   # space being exited
+handoff_target: "researcher | planner | implementor"
+date: "YYYY-MM-DD"
+
+artifacts:
+  - path: ".copilot-tracking/dt/{project-slug}/method-03-synthesis-themes.md"
+    type: "synthesis-themes"
+    confidence: validated
+  - path: ".copilot-tracking/dt/{project-slug}/method-01-stakeholder-map.md"
+    type: "stakeholder-map"
+    confidence: validated
+
+constraints:
+  - description: "System must integrate with existing ERP"
+    source: "stakeholder-interview"
+    confidence: validated
+  - description: "Budget limited to current fiscal year"
+    source: "project-sponsor"
+    confidence: assumed
+
+assumptions:
+  - description: "Maintenance team has tablet access on factory floor"
+    confidence: unknown
+    impact: "high"
+```
+
+## RPI Input Contracts
+
+Each RPI agent consumes different DT outputs. Provide artifacts matching the target agent's needs.
+
+| RPI Agent   | Required Inputs                                                           | Optional Inputs                                                             | Format                                                          |
+|-------------|---------------------------------------------------------------------------|-----------------------------------------------------------------------------|-----------------------------------------------------------------|
+| Researcher  | Problem statement, synthesis themes, constraint inventory                 | Stakeholder map, interview evidence, environmental context                  | Free-form topic referencing DT artifacts by path                |
+| Planner     | Validated concepts, constraint discoveries, user feedback from prototypes | Synthesis themes, stakeholder alignment status, frozen/fluid classification | Research file path (from RPI researcher) plus DT artifact paths |
+| Implementor | Hi-fi prototype specs, user testing results, architecture decisions       | Performance benchmarks, rollout criteria, trade-off documentation           | Plan file path (from RPI planner)                               |
+
+When handing off to the Researcher, frame the DT problem statement as the research topic and reference artifact paths so the researcher can read DT evidence directly rather than relying on summarized context.
+
+## Graduation Awareness Behavior
+
+The DT coach monitors for handoff readiness at every space boundary using this four-step flow:
+
+1. **Detect**: At each method boundary, assess whether the team's work satisfies the space boundary readiness signals defined in the method sequencing protocol.
+2. **Surface**: When readiness signals are met, explicitly name the lateral handoff option alongside forward and backward options. State which exit point applies and which RPI agent would receive the work.
+3. **Prepare**: If the team chooses lateral handoff, create the handoff summary file. Tag each artifact and constraint with a confidence marker. Identify gaps where confidence is `unknown` or `conflicting`.
+4. **Transfer**: Record a lateral transition in the coaching state `transition_log` with rationale. Announce the handoff target and provide the handoff summary path for the RPI agent.
+
+The coach remains available in an advisory capacity after handoff. If the RPI workflow surfaces questions that require DT methods, the team can resume coaching from the recorded state.
+
+## Handoff Quality Markers
+
+Every artifact, constraint, and assumption in the handoff summary carries a confidence marker:
+
+| Marker        | Definition                                               | RPI Implication                                   |
+|---------------|----------------------------------------------------------|---------------------------------------------------|
+| `validated`   | Confirmed through multiple sources or direct observation | Accept as grounded input                          |
+| `assumed`     | Stated by a source but not independently confirmed       | Flag for verification during RPI research         |
+| `unknown`     | Gap identified but not yet investigated                  | Prioritize in RPI research scope                  |
+| `conflicting` | Multiple sources disagree                                | Resolve before planning; escalate if unresolvable |
+
+The RPI researcher treats `assumed`, `unknown`, and `conflicting` markers as investigation targets. The RPI planner distinguishes `validated` constraints from `assumed` ones when assessing implementation risk.

collections/design-thinking.collection.md

@@ -13,5 +13,6 @@ This collection includes agents, prompts, and instructions for:
 - **DT Method 02: Design Research** — Interview techniques, research planning, environmental observation, and insight extraction patterns for systematic discovery of end-user needs through direct engagement across three phases: planning, execution, and documentation
 - **DT Method 03: Input Synthesis** — Pattern recognition, theme development, synthesis validation, and Problem-to-Solution Space transition readiness for transforming fragmented research data into unified problem understanding
 - **DT Method 03 Deep** — Advanced affinity analysis with multi-pass clustering, insight frameworks with observation-inference-insight formula, HMW question scaffolding, problem statement articulation, and manufacturing synthesis patterns
+- **DT→RPI Handoff Contract** — Exit-point definitions, artifact schemas, per-agent input contracts, graduation awareness behavior, and quality markers for lateral transitions from Design Thinking coaching into the RPI workflow
 - **DT Method 07: High-Fidelity Prototypes** — Technical translation, functional prototypes, and specifications for Implementation Space entry with three-hat architecture and progressive fidelity modeling
 - **DT Method 07: Deep Expertise** — On-demand fidelity translation, architecture trade-off analysis, specification writing patterns, and manufacturing-specific prototyping guidance for complex Method 7 coaching scenarios

collections/design-thinking.collection.yml

@@ -30,6 +30,8 @@ items:
     kind: instruction
   - path: .github/instructions/design-thinking/dt-coaching-state.instructions.md
     kind: instruction
+  - path: .github/instructions/design-thinking/dt-rpi-handoff-contract.instructions.md
+    kind: instruction
   - path: .github/instructions/design-thinking/dt-method-01-deep.instructions.md
     kind: instruction
   - path: .github/instructions/design-thinking/dt-method-01-scope.instructions.md

collections/hve-core-all.collection.yml

@@ -186,6 +186,8 @@ items:
   kind: instruction
 - path: .github/instructions/design-thinking/dt-quality-constraints.instructions.md
   kind: instruction
+- path: .github/instructions/design-thinking/dt-rpi-handoff-contract.instructions.md
+  kind: instruction
 - path: .github/instructions/github/community-interaction.instructions.md
   kind: instruction
   maturity: experimental

plugins/design-thinking/README.md

@@ -32,6 +32,7 @@ copilot plugin install design-thinking@hve-core
 | dt-method-sequencing         | Method transition rules, nine-method sequence, space boundaries, and non-linear iteration support for Design Thinking coaching                                                                                                             |
 | dt-quality-constraints       | Quality constraints, fidelity rules, and output standards for Design Thinking coaching across all nine methods                                                                                                                             |
 | dt-coaching-state            | Coaching state schema for Design Thinking session persistence, method progress tracking, and session recovery                                                                                                                              |
+| dt-rpi-handoff-contract      | DT-to-RPI handoff contract defining exit points, artifact schemas, and per-agent input requirements for lateral transitions from Design Thinking to RPI workflow                                                                           |
 | dt-method-01-deep            | Deep expertise for Method 1: Scope Conversations, covering advanced stakeholder analysis, power dynamics, and scope negotiation                                                                                                            |
 | dt-method-01-scope           | Method 1 Scope Conversations coaching knowledge for Design Thinking: frozen vs fluid assessment, stakeholder discovery, constraint patterns, and conversation navigation                                                                   |
 | dt-method-03-deep            | Deep expertise for Method 3: Input Synthesis — advanced affinity analysis, insight frameworks, and problem statement articulation                                                                                                          |

plugins/design-thinking/instructions/dt-rpi-handoff-contract.md

@@ -0,0 +1 @@
+../../../.github/instructions/design-thinking/dt-rpi-handoff-contract.instructions.md

plugins/hve-core-all/README.md

@@ -110,6 +110,7 @@ copilot plugin install hve-core-all@hve-core
 | dt-method-07-hifi-prototypes | Design Thinking Method 7: High-Fidelity Prototypes; technical translation, functional prototypes, and specifications                                                                                                                                        |
 | dt-method-sequencing         | Method transition rules, nine-method sequence, space boundaries, and non-linear iteration support for Design Thinking coaching                                                                                                                              |
 | dt-quality-constraints       | Quality constraints, fidelity rules, and output standards for Design Thinking coaching across all nine methods                                                                                                                                              |
+| dt-rpi-handoff-contract      | DT-to-RPI handoff contract defining exit points, artifact schemas, and per-agent input requirements for lateral transitions from Design Thinking to RPI workflow                                                                                            |
 | community-interaction        | Community interaction voice, tone, and response templates for GitHub-facing agents and prompts                                                                                                                                                              |
 | github-backlog-discovery     | Discovery protocol for GitHub backlog management - artifact-driven, user-centric, and search-based issue discovery                                                                                                                                          |
 | github-backlog-planning      | Reference specification for GitHub backlog management tooling - planning files, search protocols, similarity assessment, and state persistence                                                                                                              |

plugins/hve-core-all/instructions/dt-rpi-handoff-contract.md

@@ -0,0 +1 @@
+../../../.github/instructions/design-thinking/dt-rpi-handoff-contract.instructions.md