|
| 1 | +--- |
| 2 | +description: 'Assess DT project state and recommend next method with sequencing validation - Brought to you by microsoft/hve-core' |
| 3 | +agent: dt-coach |
| 4 | +argument-hint: "[project-slug=...]" |
| 5 | +--- |
| 6 | + |
| 7 | +# DT Method Next |
| 8 | + |
| 9 | +## Inputs |
| 10 | + |
| 11 | +* ${input:project-slug}: (Optional) Project slug identifying the DT project directory. If omitted, inferred from open files under `.copilot-tracking/dt/` or from conversation context. |
| 12 | + |
| 13 | +## Requirements |
| 14 | + |
| 15 | +### 1. Locate Project Directory |
| 16 | + |
| 17 | +**Goal:** Find the coaching state file for the specified or inferred project. |
| 18 | + |
| 19 | +* Derive project-slug from input, open files, or conversation context |
| 20 | +* Look for coaching state at `.copilot-tracking/dt/{project-slug}/coaching-state.md` |
| 21 | +* If not found and multiple projects exist, list available projects with last session dates and ask user to select |
| 22 | +* **Edge case — No project found:** If no DT project exists, respond: "No Design Thinking project found. Start a new project by running `/dt-start-project projectName='...'` with your project name." |
| 23 | + |
| 24 | +### 2. Read and Assess Current State |
| 25 | + |
| 26 | +**Goal:** Extract current method, space, completion status, and progress indicators. |
| 27 | + |
| 28 | +* Read the coaching state YAML frontmatter: |
| 29 | + * `current.method` (1-9): active method number |
| 30 | + * `current.space` (problem|solution|implementation): derived from method number |
| 31 | + * `current.phase`: free-text step within current method |
| 32 | + * `methods_completed`: array of completed method numbers |
| 33 | + * `transition_log`: history of method changes with rationales |
| 34 | + * `session_log`: recent session summaries |
| 35 | + * `artifacts`: list of generated artifacts with paths |
| 36 | +* Scan the project directory for artifact subdirectories matching `method-{NN}-*/` patterns |
| 37 | +* Assess method completeness by comparing artifacts against exit signals from `.github/instructions/design-thinking/dt-method-sequencing.instructions.md` |
| 38 | + |
| 39 | +### 3. Determine Next Method Recommendation |
| 40 | + |
| 41 | +**Goal:** Suggest appropriate next method based on state analysis and sequencing rules. |
| 42 | + |
| 43 | +Apply progression logic: |
| 44 | + |
| 45 | +* **Forward progression (primary path):** |
| 46 | + * If current method has artifacts and exit signals met → suggest method + 1 |
| 47 | + * At space boundaries (3→4, 6→7): verify readiness signals before suggesting transition |
| 48 | + |
| 49 | +* **Backward iteration (secondary path):** |
| 50 | + * If current method reveals gaps in prior work → suggest returning to earlier method with rationale |
| 51 | + * Common patterns: prototype issues → Method 2/3, brainstorming failure → Method 3, concept misalignment → Method 1 |
| 52 | + |
| 53 | +* **Lateral transitions:** |
| 54 | + * If all 9 methods complete → suggest iteration on Method 9 or handoff to RPI workflow |
| 55 | + * If user requests skipping methods → explain sequencing rationale and offer to proceed with caution |
| 56 | + |
| 57 | +* **Edge case — All complete:** If `methods_completed` includes 1-9, respond: "All 9 Design Thinking methods are complete. You can iterate on Method 9 for optimization, or hand off to RPI workflow for implementation planning. What would you like to focus on?" |
| 58 | + |
| 59 | +* **Edge case — Iteration loop detected:** If the same method or method pair appears 3+ times in the last 6 `transition_log` entries, acknowledge the iteration explicitly: "I notice you've returned to Method [N] multiple times. This suggests [observation about missing foundation]. Would you like to revisit the underlying challenge or continue refining Method [N]?" |
| 60 | + |
| 61 | +### 4. Output Format and Recommendation |
| 62 | + |
| 63 | +**Goal:** Present project status summary with clear next steps. |
| 64 | + |
| 65 | +Provide a concise summary including: |
| 66 | + |
| 67 | +* **Project:** Display name and slug |
| 68 | +* **Current Method:** Number, name, and phase description |
| 69 | +* **Progress:** Count of completed methods out of 9 |
| 70 | +* **Recent Work:** Summary from last session log entry |
| 71 | +* **Key Artifacts:** Highlight 2-3 critical artifacts from current method directory |
| 72 | + |
| 73 | +Then present recommendation: |
| 74 | + |
| 75 | +* **Suggested Next Method:** Number and name with rationale tied to exit signals or discovered gaps |
| 76 | +* **Transition Type:** Forward progression, backward iteration, or lateral handoff |
| 77 | +* **Readiness Check:** For space boundary transitions, validate these signals: |
| 78 | + * 3→4 (Problem→Solution): Themes validated across sources, team alignment confirmed, HMW questions formulated |
| 79 | + * 6→7 (Solution→Implementation): Lo-fi prototypes tested with real users, core assumptions validated, concepts narrowed to 1-2 directions |
| 80 | +* **User Choice:** "Does this direction make sense, or would you prefer to target a different method?" |
| 81 | + |
| 82 | +### 5. Delegate to DT Coach |
| 83 | + |
| 84 | +After presenting the recommendation, wait for user confirmation of the suggested method or their choice of a different method. Once confirmed, transition coaching into the target method by: |
| 85 | + |
| 86 | +* Updating `coaching-state.md` with new `current.method` value |
| 87 | +* Adding transition log entry with rationale and date |
| 88 | +* Loading the target method instruction file for method-specific knowledge |
| 89 | +* Beginning active coaching at the appropriate phase within the target method |
| 90 | + |
| 91 | +--- |
| 92 | + |
| 93 | +Assess the Design Thinking project state and recommend the next method to pursue based on completion indicators and sequencing rules. |
0 commit comments