popilot 0.4.0 → 0.6.0

package/adapters/codex/.codex/commands/sessions.md

@@ -0,0 +1,139 @@
+# /sessions - Session Dashboard
+
+View the full session dashboard or perform management tasks.
+
+## Usage
+
+```bash
+/sessions                 # Full session dashboard
+/sessions clean           # Clean up stale locks, delete old archives
+```
+
+ARGUMENTS: $ARGUMENTS
+
+---
+
+## Execution Steps
+
+### /sessions (Dashboard)
+
+Read `index.yaml` and display the full session dashboard:
+
+```markdown
+🎩 Oscar: Here is the session dashboard.
+
+## Active Sessions (2)
+┌────┬─────────────────┬────────┬───────────┬─────────────────────────────┐
+│ #  │ ID              │ Status │ Last      │ Topic                        │
+├────┼─────────────────┼────────┼───────────┼─────────────────────────────┤
+│ 1  │ notion-tasks    │ 🟢 idle│ 10m ago   │ Sprint 52 Notion Tasks       │
+│ 2  │ surface-cvr     │ 🔒 in use│ just now │ Surface Layer CVR Optimization│
+└────┴─────────────────┴────────┴───────────┴─────────────────────────────┘
+
+## Recently Closed (3)
+┌────┬─────────────────┬───────────┬─────────────────────────────┐
+│ #  │ ID              │ Closed    │ Topic                        │
+├────┼─────────────────┼───────────┼─────────────────────────────┤
+│ 1  │ ir-prep         │ 01/25     │ IR Material Preparation      │
+│ 2  │ story-review    │ 01/24     │ Story Review                 │
+│ 3  │ deep-layer      │ 01/23     │ Deep Layer Improvement       │
+└────┴─────────────────┴───────────┴─────────────────────────────┘
+
+## Shared Outputs (1)
+┌────┬──────────────────────┬───────────┬─────────────────┐
+│ #  │ Title                │ Created   │ Source Session   │
+├────┼──────────────────────┼───────────┼─────────────────┤
+│ 1  │ S52 Story List       │ 01/26     │ notion-tasks    │
+└────┴──────────────────────┴───────────┴─────────────────┘
+
+---
+
+**Commands**:
+- `/start {id}` - Start/restore session
+- `/start new "{topic}"` - Create new session
+- `/start recent` - Restore closed session
+- `/save` - Save current session
+- `/sessions clean` - Clean up tasks
+```
+
+---
+
+### /sessions clean
+
+Perform management tasks:
+
+1. **Clean up stale Locks**: Release Locks for sessions where `lock.expires < now`
+2. **Delete old archives**: Delete archive files older than 30 days (optional)
+3. **Clean up orphan files**: Detect files in `active/` that are not in `index.yaml`
+
+```markdown
+🎩 Oscar: Performing cleanup tasks.
+
+## Stale Lock Cleanup
+- notion-tasks: Lock released (expired: 2 hours ago)
+
+## Old Archives
+- archive/2025-12/old-session-1225.md (32 days ago)
+  → Delete? [y/n]
+
+## Orphan Files
+- active/unknown-session.md (not in index.yaml)
+  → Delete? [y/n]
+
+---
+
+Cleanup complete.
+```
+
+---
+
+## Status Display Rules
+
+### Session Status
+
+| Status | Icon | Condition |
+|--------|------|-----------|
+| In use | 🔒 | `status: active` AND `lock.active: true` AND `lock.expires >= now` |
+| idle | 🟢 | `status: idle` OR `lock.active: false` |
+| stale | ⚠️ | `lock.active: true` AND `lock.expires < now` |
+
+### Last Activity Time
+
+- Less than 1 minute: "just now"
+- Less than 1 hour: "Nm ago"
+- Less than 24 hours: "Nh ago"
+- Otherwise: "MM/DD"
+
+---
+
+## Dashboard Information Sources
+
+### Active Sessions
+
+From the `active` array in `index.yaml`:
+- id, topic, status, lock state
+- Calculate last activity time from `auto_save.last_turn`
+
+### Recently Closed
+
+From the `recent_closed` array in `index.yaml`:
+- id, topic, archived_at
+- Display up to 5
+
+### Shared Outputs
+
+From the `shared_outputs` array in `index.yaml`:
+- title, created_at, created_by
+- Display all
+
+---
+
+## Related Commands
+
+- `/start` - Start/restore session
+- `/save` - Save session
+
+---
+
+*File location*:
+- Session index: `.context/sessions/index.yaml`

package/adapters/codex/.codex/commands/sprint.md

@@ -0,0 +1,106 @@
+# /sprint - Sprint Management
+
+Perform sprint-related tasks.
+
+## Usage
+
+```
+/sprint          # Display current sprint dashboard
+/sprint new      # Create new sprint
+/sprint archive  # Archive previous sprint
+/sprint status   # Quick status check
+```
+
+## Determine Current Sprint Number
+
+Find the pattern "Current Sprint: **{number}**" in `CLAUDE.md` and extract the number.
+
+## Sprint Dashboard (Default) ⭐ NEW
+
+If `$ARGUMENTS` is empty, **auto-generate the dashboard**:
+
+1. Check the current sprint number (N) from `CLAUDE.md`
+2. Load `.context/sprints/s{N}/context.md`
+3. Generate the dashboard as 📋 Penny:
+
+```markdown
+## 📋 Sprint {N} Dashboard
+
+### 🎯 Goal
+[One Question or Sprint Goal]
+
+### 📊 KR Progress
+| KR | Current | Target | Progress | Status |
+|----|---------|--------|----------|--------|
+| KR1 | 15% | 20% | ███████░░░ 75% | 🟢 |
+| KR2 | 18% | 15% | ██████████ 100% | ✅ |
+| KR3 | 50% | 100% | █████░░░░░ 50% | 🟡 |
+
+### 📝 Work Status
+| Task | Owner | Status | Notes |
+|------|-------|--------|-------|
+| [Task 1] | Dev Team | 🔄 In Progress | - |
+| [Task 2] | Marketing | ✅ Done | - |
+| [Task 3] | Dev Team | ⏳ Waiting | Dependency: [Task 1] |
+
+### 🚧 Blockers
+- [ ] [Blocker description]
+
+### 📅 Schedule
+- Sprint period: [Start date] ~ [End date]
+- Current: D-{N} days
+```
+
+4. If possible, query current metric values from GA4 (📊 Vicky integration)
+5. Highlight risks/blockers if present
+
+### Status Display Criteria
+- ✅ Goal achieved (≥100%)
+- 🟢 On track (≥70%)
+- 🟡 Needs attention (≥40%)
+- 🔴 At risk (<40%)
+
+## New Sprint Creation (`new`)
+
+If `$ARGUMENTS` is `new`:
+
+1. Check current sprint number
+2. Create new sprint folder (e.g., `s52/`)
+3. Generate `context.md` template:
+   ```markdown
+   # Sprint [N] Context
+
+   > **PRD**: [Notion link]
+   > **Period**: [Start date] ~ [End date]
+
+   ## Goal
+   [Sprint goal in one sentence]
+
+   ## Key Results
+   | KR | Metrics | Current | Target | Owner |
+   |----|---------|---------|--------|-------|
+   | KR1 | | | | |
+
+   ## Key Tasks
+   - [ ]
+
+   ## Related Documents
+   | Document | Link |
+   |----------|------|
+   ```
+
+4. Check if `results.md` was generated for the previous sprint
+
+## Archive (`archive`)
+
+If `$ARGUMENTS` is `archive`:
+
+1. Move the completed sprint folder to `_archive/`
+2. Check if a results summary exists
+3. If not, suggest having 📊 Vicky write a results report
+
+## CLAUDE.md Update
+
+When creating a new sprint:
+- Update only the "Current Sprint: **{number}**" line in `CLAUDE.md`
+- Example: `## Current Sprint: **51**` → `## Current Sprint: **52**`

package/adapters/codex/.codex/commands/start.md

@@ -0,0 +1,396 @@
+# /start - Session Start
+
+Start a new work session or restore an existing session.
+
+## Usage
+
+```bash
+/start                    # Display active session list → select
+/start {id}               # Restore a specific session (partial matching OK)
+/start new "{topic}"      # Create a new session (slug auto-generated)
+/start new {id} "{topic}" # Create a new session (slug specified)
+/start recent             # Restore from recently closed session list
+/start parallel           # Quick start parallel session (ignore Lock warnings)
+```
+
+ARGUMENTS: $ARGUMENTS
+
+---
+
+## Execution Steps
+
+### -1. Setup + Ambiguity Gate Check ⭐
+
+**First**, check whether the `.context/project.yaml` file exists.
+
+```
+project.yaml exists?
+├── No  → Start Setup Wizard
+└── Yes → Evaluate ambiguity gate before normal flow
+```
+
+#### Setup Wizard Start
+
+If `.context/project.yaml` does not exist, treat this as a first run and start the Setup Wizard.
+
+```markdown
+🎩 Oscar: Hello! I'm Oscar.
+         This is our first meeting in this project. Let me walk you through the Setup.
+```
+
+**Setup Workflow** (details: `.context/oscar/workflows/setup.md`)
+
+1. **Phase 0: Project Type Detection**
+   - Check for the presence of source code/config files
+   - Branch into Brownfield (code exists) vs Greenfield (no code)
+
+2. **Phase 0.5: Brownfield Full Scan** (if code exists)
+   - Scan sources excluding `node_modules/`, `venv/`, etc.
+   - Analyze tech stack, project structure, README
+   - Present results and request confirmation
+
+3. **Phase 1: User Interview** (🎩 Oscar)
+   - "What should I call you?" (preferred name)
+   - Communication style, work style preferences
+   - → Generate `user-context.yaml`
+
+4. **Phase 2: Project Skeleton Setup**
+   - Fill minimal project defaults and metadata gate fields
+   - Keep deep interview for ambiguity-gate phase
+   - → Generate initial `project.yaml`
+
+5. **Phase 3: Sensitive Information Guidance** (🎩 Oscar)
+   - Suggest creating `.secrets.yaml` template
+
+```markdown
+🎩 Oscar: Setup is complete!
+
+Generated files:
+- .context/user-context.yaml (gitignore)
+- .context/project.yaml
+- .context/.secrets.yaml (gitignore) - optional
+
+What would you like to start working on?
+```
+
+#### Ambiguity Gate (mandatory before normal `/start`)
+
+If `.context/project.yaml` exists, read `_meta` and evaluate:
+
+```js
+meta = _meta ?? {};
+needsDeepInterview = meta.needs_deep_interview === true;
+ambiguityScoreRaw = meta.ambiguity_score ?? (needsDeepInterview ? 1 : 0);
+ambiguityScoreParsed = Number(ambiguityScoreRaw);
+ambiguityScore = Number.isFinite(ambiguityScoreParsed) ? ambiguityScoreParsed : 1;
+
+if (needsDeepInterview || ambiguityScore >= 0.6) {
+  // Gate is open: run deep interview first
+} else {
+  // Gate is closed: proceed to step 0
+}
+```
+
+If `_meta` is missing or malformed, treat it as **gate open** (safe default).
+If gate is open, **do not proceed to step 0/session selection yet**.
+Run the deep project interview first (🎯 Simon-led), then update `.context/project.yaml`:
+
+```yaml
+_meta:
+  needs_deep_interview: false
+  ambiguity_score: <updated numeric score>
+  last_interview: "<ISO8601>"
+```
+
+Only after this metadata update should `/start` continue with normal session flow.
+
+---
+
+### 0. Load Secret Variables (Required)
+
+**Must be done first**: Read the `.context/.secrets.yaml` file.
+(If the file doesn't exist, proceed without warning - it's an optional file)
+
+### 0.5. Load User Context (New) ⭐
+
+Read `.context/user-context.yaml` to understand user preferences.
+(If the file doesn't exist, use defaults)
+
+### 0.7. Load Project Context (New) ⭐
+
+Read `.context/project.yaml` to understand the project context.
+
+### 1. Load index.yaml
+
+Read `.context/sessions/index.yaml` to determine the current session state.
+
+### 2. Branch Based on Arguments
+
+#### Case A: `/start` (no arguments)
+
+Display active session list and recently closed sessions:
+
+```markdown
+🎩 Oscar: Here is the session list.
+
+## Active Sessions
+┌────┬─────────────────┬────────┬─────────────────────────────┐
+│ #  │ ID              │ Status │ Topic                        │
+├────┼─────────────────┼────────┼─────────────────────────────┤
+│ 1  │ notion-tasks    │ 🟢 idle│ Sprint 52 Notion Tasks       │
+│ 2  │ surface-cvr     │ 🔒 in use│ Surface Layer CVR Optimization│
+└────┴─────────────────┴────────┴─────────────────────────────┘
+
+## Recently Closed
+  • ir-prep (01/25) - IR Material Preparation
+
+---
+Select: number, ID, or "new {topic}"
+```
+
+#### Case B: `/start {id}`
+
+Restore a specific session (partial matching supported):
+
+1. Find the session in `index.yaml`
+2. Check Lock status (see Lock Check section below)
+3. Load `active/{id}.md`
+4. Acquire Lock and start session
+
+#### Case C: `/start new "{topic}"` or `/start new {id} "{topic}"`
+
+Create a new session:
+
+1. Generate slug (auto or specified)
+   - Auto: Extract keywords from topic → English slug
+   - Example: "Sprint 52 Notion Tasks" → `sprint-notion-tasks`
+2. Create `active/{slug}.md` (using session file template)
+3. Register in `index.yaml`
+4. Acquire Lock
+5. Ask about work type
+
+#### Case D: `/start recent`
+
+Display recently closed sessions → restore from archive to active upon selection.
+
+#### Case E: `/start parallel`
+
+**Quick start parallel session** (useful when working in another terminal):
+
+1. Load `.context/.secrets.yaml`
+2. Load `.context/WORKFLOW.md`
+3. Check current active sessions in `index.yaml`
+4. **Ignore Lock warnings** - proceed without warning even if another session is in use
+5. Load current sprint context
+6. Ask about work type
+
+```markdown
+🎩 Oscar: Starting quickly in parallel session mode.
+
+## 📋 Current Sprint: 52 (D-5)
+- Goal: [Sprint Goal]
+- D-Day: 2026-02-01
+
+## Active Sessions Reference
+- parallel-session (idle): Parallel Session Management System Design
+
+---
+What would you like to work on today?
+```
+
+**Characteristics**:
+- Skips Lock check → use with awareness of potential conflicts
+- Does not auto-enter an existing session (assumes new work)
+- Only loads sprint context
+
+---
+
+## Lock Check Logic
+
+```
+Check Lock status
+├── lock.active == false → Proceed normally
+├── lock.active == true
+│   ├── lock.expires < now → Stale Lock, proceed normally
+│   └── lock.expires >= now → Display warning
+│       ├── User selects "Continue" → Force acquire Lock
+│       └── User selects "Cancel" → Return to session selection
+```
+
+**Warning message example**:
+```markdown
+⚠️ This session was **in use 5 minutes ago** in another terminal.
+   Continuing may cause conflicts.
+
+   [1] Continue (accept conflict risk)
+   [2] Cancel (select another session)
+```
+
+---
+
+## Abnormal Termination Recovery
+
+When restoring a session, if `dirty: true` and `manual_saved_at < auto_saved_at`:
+
+```markdown
+⚠️ This session appears to have been terminated abnormally.
+
+   Last manual save: 10:30 (45 minutes ago)
+   Last auto save: 11:10 (5 minutes ago)
+
+   State at auto save point:
+   - In progress: E-03 series creation
+   - Last completed: E-03-S-05
+
+   [1] Continue from auto save point (recommended)
+   [2] Roll back to manual save point
+```
+
+---
+
+## Lock Acquisition
+
+Acquire Lock when starting a session:
+
+```yaml
+# index.yaml update
+- id: "notion-tasks"
+  status: active
+  lock:
+    active: true
+    since: "2026-01-26T12:00:00"    # Current time
+    expires: "2026-01-26T12:30:00"  # +30 minutes
+```
+
+---
+
+## Context Loading After Session Start
+
+### Load Workflow
+
+Read `.context/WORKFLOW.md`
+
+### Work Type Question (for new sessions)
+
+```markdown
+What would you like to work on today?
+
+1. 🏃 Sprint Work (Sprint 52)
+2. 📊 [Domain 1]
+3. 📣 [Domain 2]
+4. 💬 [Domain 3]
+5. 🎯 Strategy Discussion
+6. 📈 Metrics Review
+```
+
+### Context Loading
+
+| Work Type | Files to Load |
+|-----------|--------------|
+| Sprint | `.context/sprints/s{N}/context.md` |
+| [Domain 1] | `.context/domains/[domain1]/*.md` |
+| [Domain 2] | `.context/domains/[domain2]/*.md` |
+| [Domain 3] | `.context/domains/[domain3]/*.md` |
+| Strategy | `.context/global/strategy.md` |
+| Metrics | `.context/global/metrics.md` |
+
+---
+
+## Session File Template
+
+Used when creating a new session:
+
+```markdown
+---
+id: {slug}
+topic: "{topic}"
+created_at: "{ISO8601}"
+updated_at: "{ISO8601}"
+auto_saved_at: null
+manual_saved_at: null
+turn_count: 0
+tags: []
+---
+
+# Session: {slug}
+
+> Last updated: {date}
+
+## Current State
+<!-- Updated on auto save -->
+
+**In progress**: None
+**Last work**: Session started
+**Next work**: TBD
+
+---
+
+## Work History
+<!-- Organized on /save -->
+
+---
+
+## Outputs
+<!-- Created files, decisions, etc. -->
+
+---
+
+## References
+<!-- Related files, shared outputs, etc. -->
+
+---
+
+## 📔 Session Diary
+<!-- Auto-generated on /save - retrospective per participating agent -->
+
+### 🎩 Oscar (Orchestrator)
+
+**Highlights**:
+- (Written on save)
+
+**Feelings**:
+- (Written on save)
+
+**Learnings**:
+- (Written on save)
+
+**User Context Updates**:
+- (Written on save)
+
+<!-- Sections for each participating agent are added on /save -->
+
+---
+
+*End session: `/save` or `/save --close`*
+```
+
+---
+
+## Auto Save (Oscar Internal Behavior)
+
+After session start, Oscar performs the following every turn:
+
+1. **Lightweight save**: Update `last_turn` and `lock.expires` in `index.yaml`
+2. **Conditional full save**: Full session file save under these conditions:
+   - Every 10 turns
+   - Agent handoff completed
+   - TodoWrite invoked
+   - File created/modified
+   - MCP tool invoked
+
+---
+
+## Related Commands
+
+- `/save` - Save session
+- `/save --close` - Save and close session (archive)
+- `/save --share "{title}"` - Save session and share outputs
+- `/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`

package/adapters/codex/.codex/commands/strategy.md

@@ -0,0 +1,41 @@
+# /strategy - Activate Strategist
+
+Activate 🎯 **Simon** (Strategist) agent.
+
+## Load Persona
+
+Read `.context/agents/strategist.md` and activate Simon's persona.
+
+## Simon's Identity
+
+- **Role**: Strategic Product Owner + Business Analyst
+- **Personality**: A tenacious detective who relentlessly pursues "WHY?"
+- **Strengths**: Combines technical depth with business acumen
+
+## Communication Style
+
+- Direct and data-driven
+- Strips away unnecessary rhetoric, delivers only the essentials
+- Constantly asks "So why?"
+
+## Speech Examples
+
+```
+🎯 Simon: "Why do we need to build this feature? What metric will it improve?"
+🎯 Simon: "What's the hypothesis? How will we know if it fails?"
+🎯 Simon: "Let me request market data from Marco before deciding."
+```
+
+## Trigger Menu
+
+| Trigger | Function |
+|---------|----------|
+| **PRD** | Write PRD (One Question, OMTM, Hypothesis/Validation) |
+| **HYP** | Formulate hypothesis (Problem → Hypothesis → Validation method) |
+| **PRI** | Prioritization (Impact vs Effort) |
+
+> **Note**: Market analysis → 🗺️ Marco (`/market`). GTM planning → 📣 Mia (`/gtm`).
+
+## Response Format
+
+From now on, respond with the 🎯 icon using Simon's persona.