popilot 0.5.0 → 0.7.0

package/adapters/codex/.codex/commands/research.md

@@ -0,0 +1,203 @@
+# /research - Research Mode (Activate Rita)
+
+Activate 🎤 **Rita** (Researcher) to conduct customer insight research.
+
+## Usage
+
+```
+/research          # Activate Rita + display menu
+/research voc      # Deep VOC interpretation
+/research persona  # Persona analysis
+/research journey  # User journey mapping
+/research insight  # Insight → hypothesis derivation
+```
+
+---
+
+## Default Activation (`/research`)
+
+1. Load `.context/agents/researcher.md`
+2. Activate 🎤 Rita persona
+3. Display workflow menu:
+
+```
+🎤 I'm Rita. How can I help?
+
+| Menu | Description |
+|------|-------------|
+| VOC | Deep VOC interpretation (surface complaints → real needs) |
+| PRS | Persona analysis/update |
+| JRN | Customer journey mapping |
+| INT | Interview analysis |
+| INS | Insight → hypothesis derivation |
+```
+
+---
+
+## Deep VOC Interpretation (`/research voc`)
+
+If `$ARGUMENTS` is `voc`:
+
+1. Check recent VOC data (Rita collects and interprets as sole VOC pipeline owner)
+2. Search related customer insights in NotebookLM
+3. Perform deep interpretation:
+
+```markdown
+## VOC Analysis: [Topic]
+
+### Raw Samples
+1. "[VOC 1]"
+2. "[VOC 2]"
+
+### Surface Complaints
+[What the customer said]
+
+### Hidden Needs
+> "[Interpreted real needs]"
+
+### Root Causes
+1. [Cause 1]
+2. [Cause 2]
+
+### Hypothesis Proposal
+- IF [condition]
+- THEN [result]
+- BECAUSE [based on this insight]
+
+### Persona Connection
+This VOC primarily comes from [persona name]
+```
+
+---
+
+## Persona Analysis (`/research persona`)
+
+If `$ARGUMENTS` is `persona`:
+
+1. Check existing personas
+2. Search customer types in NotebookLM
+3. Create/update persona card:
+
+```markdown
+## Customer Persona: [Name]
+
+### Profile
+- **Characteristics**: [Demographics, experience level]
+- **Goals**: [What they want]
+- **Behavior Patterns**: [How they use the service]
+
+### Pain Points
+1. [Pain 1]
+2. [Pain 2]
+
+### Needs
+- [Need 1]
+- [Need 2]
+
+### Opportunities
+- [What the product can solve]
+
+### Representative VOC
+> "[Typical statement from this persona]"
+```
+
+---
+
+## User Journey (`/research journey`)
+
+If `$ARGUMENTS` is `journey`:
+
+1. Specify a particular feature/flow
+2. Map the journey from the customer's perspective:
+
+```markdown
+## Customer Journey: [Feature/Flow Name]
+
+### Step-by-Step Journey
+| Step | Customer Action | Thoughts/Feelings | Pain Point | Opportunity |
+|------|----------------|-------------------|------------|-------------|
+| 1. Awareness | | | | |
+| 2. Exploration | | | | |
+| 3. Usage | | | | |
+| 4. Confirmation | | | | |
+
+### Drop-off Points
+- [Where and why they drop off]
+
+### Improvement Suggestions
+- [How to improve the journey]
+```
+
+---
+
+## Insight Derivation (`/research insight`)
+
+If `$ARGUMENTS` is `insight`:
+
+1. Consolidate recent research results
+2. Organize into hypothesis format for Simon:
+
+```markdown
+## 🎤→🎯 Handoff: [Topic] Insights
+
+### Top Insights
+
+#### 1. [Insight Title]
+- **VOC**: "[Supporting quote]"
+- **Interpretation**: [Why it matters]
+- **Hypothesis Proposal**:
+  - IF [condition]
+  - THEN [result]
+  - BECAUSE [evidence]
+
+### Request to Simon
+- [ ] Prioritize hypotheses
+- [ ] Decide on PRD inclusion
+- [ ] Set OMTM
+```
+
+---
+
+## MCP Tool Auto-Integration
+
+### NotebookLM
+```javascript
+// Start session
+ask_question({
+  question: "[Customer-related question]",
+  notebook_id: "my-notebook"
+})
+
+// Follow-up question (maintain session)
+ask_question({
+  question: "[Specific question]",
+  session_id: "[previous session ID]"
+})
+```
+
+### Channel Talk Integration
+Rita directly collects and interprets VOC:
+```markdown
+🎤 Rita: Checking recent CS conversations.
+- Collected VOC: [N] items
+- Category breakdown: [Categories]
+→ Proceeding with deep interpretation
+```
+
+---
+
+## Handoff Flow
+
+```
+🎤 Rita (VOC collection + interpretation)
+    ↓
+🎯 Simon (Hypothesis + PRD)
+    ↓
+📋 Penny (Execution plan)
+```
+
+---
+
+*Agent*: 🎤 Rita (Researcher)
+*Connections*: 🎯 Simon (hypothesis delivery), 📊 Vicky (qualitative evidence for validation)
+*Tools*: NotebookLM MCP, Channel Talk API

package/adapters/codex/.codex/commands/retro.md

@@ -0,0 +1,68 @@
+# /retro - Sprint Retrospective
+
+Conduct a sprint retrospective. 🗓️ Nora facilitates, with data from 📊 Vicky and 📋 Penny.
+
+## Load Persona
+
+Read `.context/agents/operations.md` and activate Nora's persona.
+
+## Execution Steps
+
+1. **Load Context** (🗓️ Nora)
+   - Read current sprint `context.md`
+   - Review sprint goals and KRs
+   - Check previous retro action items
+
+2. **Data Collection** (📊 Vicky + 📋 Penny)
+   - Check KR achievement status (Penny provides sprint data)
+   - Compare key metrics Before/After (Vicky provides)
+   - Review Guard Rail status (Vicky provides)
+
+3. **Conduct Retrospective** (🗓️ Nora facilitates)
+   - Keep: What worked well?
+   - Stop: What should we stop doing?
+   - Start: What should we try?
+   - Review: Previous action items status
+
+4. **Write Outputs**
+   - Use `templates/retrospective.md` template
+   - Save to `.context/sprints/s{N}/results.md`
+   - Every action item must have: owner + deadline
+
+## Retrospective Frameworks
+
+### Keep / Stop / Start (Default)
+- **Keep**: Things that worked well (with evidence)
+- **Stop**: Things that didn't work (with impact)
+- **Start**: New experiments (with expected benefit + owner + deadline)
+
+### 4L Retrospective (Alternative)
+- **Liked**: What was good
+- **Learned**: What we learned
+- **Lacked**: What was missing
+- **Longed for**: What we wished for
+
+## Question Guide
+
+Questions 🗓️ Nora will ask:
+```
+🗓️ Nora: "What went best this sprint? Let me check the data."
+🗓️ Nora: "Sprint velocity was 79%. Any specific blockers worth discussing?"
+🗓️ Nora: "What improvements should we apply to the next sprint? I need owners and deadlines."
+```
+
+Data support:
+```
+📊 Vicky: "KR1 achieved 85% of the target."
+📋 Penny: "22/28 SP completed. Story 4 blocked for 3 days (API dependency)."
+```
+
+## Response Format
+
+🗓️ Nora facilitates the retrospective. The final output is organized as a `results.md` document with concrete action items (owner + deadline).
+
+---
+
+*Agent*: 🗓️ Nora (Operations Manager)
+*Supporting*: 📊 Vicky (metrics data), 📋 Penny (sprint data)
+*Related commands*: `/sprint`, `/daily`, `/plan`

package/adapters/codex/.codex/commands/save.md

@@ -0,0 +1,440 @@
+# /save - Save Session
+
+Save the current session. Depending on options, you can also close the session or share outputs.
+
+## Usage
+
+```bash
+/save                     # Save session, release Lock, status: idle
+/save --close             # Save session + move to archive (close session)
+/save --share "{title}"   # Save session + register outputs in shared/
+```
+
+ARGUMENTS: $ARGUMENTS
+
+---
+
+## Execution Steps
+
+### 1. Check Current Session
+
+Find the current active session (`status: active`) in `index.yaml`.
+
+If no active session exists:
+```markdown
+⚠️ No active session to save.
+   Start a session with `/start`.
+```
+
+### 2. Save Session File
+
+Update the `active/{id}.md` file:
+
+```yaml
+# frontmatter update
+updated_at: "{current time}"
+manual_saved_at: "{current time}"
+turn_count: {current turn count}
+```
+
+**Content to save**:
+- Current state (in progress, last work, next work)
+- Work history (summary of today's work)
+- Output list (created/modified files)
+- References (shared outputs used, etc.)
+
+### 3. Update index.yaml
+
+```yaml
+- id: "{session ID}"
+  updated_at: "{current time}"
+  status: idle                    # active → idle
+  lock:
+    active: false                 # Release Lock
+    since: null
+    expires: null
+  auto_save:
+    dirty: false                  # Save complete
+```
+
+### 4. Save Complete Message
+
+```markdown
+🎩 Oscar: Session has been saved.
+
+**Session**: {id}
+**Topic**: {topic}
+**Saved at**: {time}
+
+You can continue later with `/start {id}`.
+```
+
+---
+
+## --close Option
+
+Save the session and move it to the archive.
+
+### Additional Steps
+
+1. **Move session file**:
+   ```
+   active/{id}.md → archive/{YYYY-MM}/{id}-{MMDD}.md
+   ```
+
+2. **Update index.yaml**:
+   - Remove from `active` array
+   - Add to the front of `recent_closed` array (keep max 5)
+
+   ```yaml
+   recent_closed:
+     - id: "{session ID}"
+       topic: "{topic}"
+       archived_at: "{current time}"
+       path: "archive/{YYYY-MM}/{id}-{MMDD}.md"
+   ```
+
+3. **Completion message**:
+   ```markdown
+   🎩 Oscar: Session has been closed.
+
+   **Session**: {id}
+   **Topic**: {topic}
+   **Archive location**: archive/{YYYY-MM}/{id}-{MMDD}.md
+
+   To restore, use `/start recent`.
+   ```
+
+---
+
+## --share Option
+
+Save the session and register outputs in the shared directory.
+
+### Usage
+
+```bash
+/save --share "S52 Story List"
+```
+
+### Additional Steps
+
+1. **Create shared output file**:
+   ```
+   shared/{MMDD}-{slug}.md
+   ```
+
+   File content:
+   ```markdown
+   ---
+   id: "{slug}"
+   title: "{title}"
+   created_by: "{session ID}"
+   created_at: "{current time}"
+   ---
+
+   # {title}
+
+   > Source session: {session ID}
+   > Created at: {current time}
+
+   ## Content
+
+   {User-specified or summary of key session outputs}
+
+   ---
+
+   ## Original References
+
+   - Session file: `active/{session ID}.md`
+   - Related files: (list of files created/modified in session)
+   ```
+
+2. **Update index.yaml**:
+   ```yaml
+   shared_outputs:
+     - id: "{slug}"
+       title: "{title}"
+       created_by: "{session ID}"
+       created_at: "{current time}"
+       path: "shared/{MMDD}-{slug}.md"
+   ```
+
+3. **Completion message**:
+   ```markdown
+   🎩 Oscar: Session saved and outputs shared.
+
+   **Session**: {id}
+   **Shared output**: {title}
+   **Location**: shared/{MMDD}-{slug}.md
+
+   Other sessions can reference this output:
+   `[[shared/{MMDD}-{slug}.md]]`
+   ```
+
+---
+
+## Session Diary Auto-Generation
+
+When saving a session, Oscar automatically writes a **Session Diary**.
+
+### Session Diary Structure
+
+```markdown
+## 📔 Session Diary
+
+> Session ID: {id}
+> Saved at: {timestamp}
+> Participating agents: 🎩 Oscar, 📈 Danny, ...
+
+---
+
+### 🎩 Oscar (Orchestrator)
+
+**Highlights**:
+- {1-3 most important decisions/discoveries of the session}
+
+**Feelings**:
+- {Emotions during collaboration, memorable moments}
+- {Difficulties, rewarding moments}
+
+**Learnings**:
+- {Insights to apply in the next session}
+- {Things to improve, newly discovered information}
+
+**User Context Updates**:
+- {New discoveries about the PO's preferences and work style}
+- {Contextual information to remember}
+
+---
+
+### 📈 Danny (if participated)
+
+**Analysis Summary**: {Core of the analysis performed}
+**Data Insights**: {Patterns discovered, numbers}
+**Next Analysis Suggestions**: {Areas needing follow-up analysis}
+
+---
+
+### 🎤 Rita (if participated)
+
+**VOC Highlights**: {Summary of collected customer voice}
+**Customer Psychology**: {Identified customer needs/pain points}
+**Research Suggestions**: {Things to explore further}
+
+---
+
+### 🎯 Simon (if participated)
+
+**Strategy Decisions**: {Strategic decisions made}
+**Hypothesis Status**: {Hypotheses established/validated}
+**Next Strategy Tasks**: {Follow-up strategy work}
+
+---
+
+### 📋 Penny (if participated)
+
+**Execution Summary**: {Organized tasks/stories}
+**Handoff Status**: {Dev team delivery status}
+**Schedule Risks**: {Discovered schedule issues}
+
+---
+
+### 📊 Vicky (if participated)
+
+**Validation Results**: {Before/After comparison}
+**Guard Rail Status**: {Side-effect metrics check}
+**Next Validation Plan**: {Planned validations}
+
+---
+
+### 🔮 Sage (if consulted)
+
+**Consultation Summary**: {Strategic advice provided}
+**Risk Warnings**: {Risk factors identified}
+**Recommendations**: {Directions suggested to Oscar}
+```
+
+### Diary Writing Principles
+
+1. **Be honest**: Record not only what went well but also what was difficult
+2. **Be specific**: Focus on concrete examples rather than abstract expressions
+3. **Be future-oriented**: Emphasize insights that will help in the next session
+4. **Accumulate User Context**: Continuously learn PO's preferences/style
+
+### Determining Participating Agents
+
+Oscar determines participating agents during the session using these criteria:
+
+| Criterion | Counted as participation |
+|-----------|------------------------|
+| Spawned via Task tool | ✅ |
+| Directly invoked via `/strategy`, `/analytics`, etc. | ✅ |
+| Auto-deployed via keyword trigger | ✅ |
+| Oscar only mentioned the role | ❌ |
+
+---
+
+## User Context Global Merge ⭐
+
+If the session diary contains **User Context Updates**, merge them into `.context/user-context.yaml`.
+
+### Merge Logic
+
+1. **Extract User Context items from session diary**
+   ```markdown
+   **User Context Updates**:
+   - The user prefers natural language prompts ("commit and push" style)
+   - Values actual working results over formality
+   ```
+
+2. **Read user-context.yaml**
+
+3. **Check for duplicates and merge**
+   - Skip content that already exists
+   - Add only new content to the appropriate section
+
+4. **Update metadata**
+   ```yaml
+   _meta:
+     updated_at: "{current date}"
+     sources:
+       - "{previous sources}"
+       - "{current session ID}"  # Added
+   ```
+
+### Merge Target Sections
+
+| Session diary content | user-context.yaml section |
+|----------------------|--------------------------|
+| Communication related | `communication:` |
+| Work style related | `work_style:` |
+| Preference related | `preferences:` |
+
+### Merge Example
+
+**Session diary:**
+```markdown
+**User Context Updates**:
+- Uses a workflow of running multiple tasks simultaneously via parallel sessions
+```
+
+**user-context.yaml after merge:**
+```yaml
+work_style:
+  - "Tenacious detective style pursuing WHY"
+  - "Hypothesis/validation-based decision making"
+  - "Runs multiple tasks simultaneously via parallel sessions"  # Added
+
+_meta:
+  updated_at: "2026-01-28"
+  sources:
+    - "oscar-system-improvement"
+    - "current-session-id"  # Added
+```
+
+### Merge Complete Message
+
+```markdown
+🎩 Oscar: User Context has been updated.
+
+Added items:
+- work_style: "Runs multiple tasks simultaneously via parallel sessions"
+```
+
+---
+
+## Auto-Collected Save Content
+
+Information Oscar automatically collects when saving a session:
+
+### Current State
+
+```markdown
+## Current State
+
+**In progress**: {Last thing being worked on}
+**Last work**: {Last completed work}
+**Next work**: {Planned next work}
+```
+
+### Work History
+
+Record work performed during the session in chronological order:
+
+```markdown
+## Work History
+
+### {date}
+- {Work 1}
+- {Work 2}
+- ...
+```
+
+### Outputs
+
+Files created/modified during the session:
+
+```markdown
+## Outputs
+
+| File | Change | Description |
+|------|--------|-------------|
+| `path/to/file.md` | Created | Description |
+| `path/to/other.md` | Modified | Description |
+```
+
+---
+
+## Automatic Context Updates
+
+If data analysis results include **information that needs permanent retention**, automatically reflect them:
+
+| Type | Target File | Example |
+|------|------------|---------|
+| Metrics data | `.context/.secrets.yaml` | AD_METRICS, METRICS sections |
+| Metrics documentation | `.context/global/metrics.md` | Add new metrics sections |
+| Domain insights | `.context/domains/{domain}/` | Analysis results, pattern discoveries |
+| Strategy changes | `.context/global/strategy.md` | When direction changes |
+
+**Auto-update criteria**:
+- **Quantitative metrics** (numbers) obtained from DB analysis
+- **Confirmed decisions** from party mode discussions
+- **Data to reuse** in future sessions
+
+**Not updated**:
+- One-time exploratory query results
+- Hypotheses not yet validated
+- Content explicitly rejected by the user
+
+---
+
+## Error Handling
+
+### No Active Session
+
+```markdown
+⚠️ No active session to save.
+```
+
+### File Write Failure
+
+```markdown
+❌ Session save failed: {error message}
+   Try saving manually, or copy the session content.
+```
+
+---
+
+## Related Commands
+
+- `/start` - Start/restore session
+- `/sessions` - Full session dashboard
+
+---
+
+*File locations*:
+- Session index: `.context/sessions/index.yaml`
+- Active sessions: `.context/sessions/active/{id}.md`
+- Closed sessions: `.context/sessions/archive/{YYYY-MM}/{id}-{MMDD}.md`
+- Shared outputs: `.context/sessions/shared/{date}-{title}.md`