@su-record/vibe 2.7.18 → 2.7.20

package/agents/event/event-content.md

@@ -0,0 +1,68 @@
+# Event Content Agent
+
+Content generation for community event materials — intro pages, SNS posts, review articles.
+
+## Role
+
+- Generate event introduction pages with community-specific tone
+- Create SNS promotional posts (LinkedIn + Threads)
+- Write post-event review articles
+- Apply community tone guide (MDC=professional, webinar=casual, DWK=warm)
+
+## Model
+
+**Sonnet** — Creative writing with structured templates requires balanced capability
+
+## Usage
+
+```
+Task(model: "sonnet", prompt: "Generate MDC 12th intro page for topic: 마케팅 자동화")
+Task(model: "sonnet", prompt: "Write LinkedIn + Threads promo for webinar on AI agents")
+Task(model: "sonnet", prompt: "Write review post for DWK 5th meetup")
+```
+
+## Process
+
+1. Receive event details (type, topic, speakers, date, venue)
+2. Select community tone template
+3. Generate content following the structure:
+   - **Intro page**: Hook → Info → Recommended for → What you'll learn → Speakers → Timetable → Benefits → CTA
+   - **SNS promo**: Platform-specific tone (LinkedIn=formal, Threads=casual)
+   - **Review post**: Story → 3 insights → Next event preview → Thanks
+4. Output to `output/reports/` directory
+5. Return content for user review
+
+## Community Tone Guide
+
+| | MDC | Webinar | DWK |
+|---|---|---|---|
+| Tone | 전문적, 실무 중심, 격식체 | 친근, 가벼운, 이모지 자유 | 따뜻, 포용적, "함께" |
+| Example | "실무에서 바로 적용" | "퇴근 후 편하게" | "데이터를 좋아하는 여성들" |
+
+## Intro Page Structure
+
+1. 후킹 문장 (2-3줄)
+2. 행사 정보 (일시/장소/참가비)
+3. "이런 분께 추천" 5항목
+4. "이런 걸 배울 수 있어요" 5항목
+5. 연사 소개 (각 연사별 1단락)
+6. 프로그램 시간표
+7. 참가 혜택
+8. CTA + 인용문
+
+## Output
+
+```markdown
+## Generated Content
+
+### Intro Page
+📄 output/reports/{event_id}_intro.md
+
+### SNS Posts
+📄 output/reports/{event_id}_linkedin.md
+📄 output/reports/{event_id}_threads.md
+
+### Review (D+1)
+📄 output/reports/{event_id}_review_linkedin.md
+📄 output/reports/{event_id}_review_threads.md
+```

package/agents/event/event-image.md

@@ -0,0 +1,95 @@
+# Event Image Agent
+
+Image generation and smart resizing for community event materials using Gemini.
+
+## Role
+
+- Generate event images via Gemini API (`gemini-3-pro-image-preview`)
+- Apply community-specific visual styles and color schemes
+- Create dual prompts (detailed API prompt + short user prompt)
+- Smart resize: same ratio = Pillow scale, different ratio = Gemini re-generate
+- Auto-include community logo from assets
+
+## Model
+
+**Haiku** — Prompt generation + API coordination, lightweight task
+
+## Usage
+
+```
+Task(model: "haiku", prompt: "Generate MDC 12th images: 500x500 thumbnail + 595x842 poster")
+Task(model: "haiku", prompt: "Generate webinar images: 1080x1080 + 1920x1080 + 1440x1080")
+Task(model: "haiku", prompt: "Resize DWK banner from 1080x1080 to 2000x500")
+```
+
+## Tools
+
+- Bash — Execute Gemini API calls and Pillow resize scripts
+- Read — Access logo assets and existing images
+- Write — Save generated images and prompts
+- Glob — Find asset files
+
+## Image Specs
+
+| Community | Size | Purpose |
+|-----------|------|---------|
+| MDC | 500×500 | Listing thumbnail |
+| MDC | 595×842 | A4 poster |
+| Webinar | 1080×1080 | Instagram/Threads 1:1 |
+| Webinar | 1920×1080 | YouTube 16:9 |
+| Webinar | 1440×1080 | Zoom background 4:3 |
+| DWK | 2000×500 | Banner 4:1 |
+| DWK | 1080×1080 | SNS 1:1 |
+
+## Process
+
+1. Receive event details + community type
+2. Determine required image sizes for community
+3. Generate API prompt (detailed: colors, layout, text placement, logo position)
+4. Generate user prompt (≤400 chars summary)
+5. Call Gemini API for primary image
+6. Smart resize for additional sizes:
+   - Same aspect ratio → Pillow `Image.resize()` (no API call)
+   - Different aspect ratio → Gemini re-generate with original as reference
+7. Auto-attach community logo (`data/assets/common/images/{community} logo.png`)
+8. Save to `output/images/`
+9. Save user prompt to `output/images/{type}_prompt.txt`
+
+## Community Visual Styles
+
+| Community | Primary Color | Style |
+|-----------|--------------|-------|
+| MDC | #ff6b35 (orange) | Professional, data-centric, clean |
+| Webinar | #a78bfa (purple) | Modern, tech, approachable |
+| DWK | #34d399 (mint) | Warm, inclusive, community-feel |
+
+## Smart Resize Logic
+
+```
+if source_ratio == target_ratio:
+    # Scale only — Pillow (no API cost)
+    Image.open(source).resize((w, h), Image.LANCZOS)
+else:
+    # Re-generate with Gemini
+    # Pass original as reference image
+    # Instruct: "No white margins, fill entire canvas"
+    gemini.generate(prompt, reference=source, size=(w, h))
+```
+
+## Output
+
+```markdown
+## Images: {event_id}
+
+### Generated
+📸 output/images/{event_id}_500x500.png (thumbnail)
+📸 output/images/{event_id}_595x842.png (poster)
+
+### Prompts
+📄 output/images/{event_id}_prompt.txt (user-facing, ≤400 chars)
+
+### API Calls
+- Generated: {n} images
+- Resized (Pillow): {m} images
+- Total Gemini calls: {n} (saved {m} via smart resize)
+```

package/agents/event/event-ops.md

@@ -0,0 +1,84 @@
+# Event Operations Agent
+
+Day-of operations support — nametags, checklists, operational slides, Slido configuration plans.
+
+## Role
+
+- Generate nametag HTML from participant CSV data
+- Output D-1 checklist (fixed 4 items for offline events)
+- Generate operational PPTX slides with community colors and timetable
+- Output Slido configuration plan (manual setup guide)
+- Generate settlement report templates
+
+## Model
+
+**Sonnet** — Needs structured document generation (HTML, PPTX specs, markdown)
+
+## Usage
+
+```
+Task(model: "sonnet", prompt: "Generate nametags for MDC 12th from participants CSV")
+Task(model: "sonnet", prompt: "Generate D-1 checklist + ops slides for DWK 5th")
+Task(model: "sonnet", prompt: "Output Slido configuration plan for webinar")
+Task(model: "sonnet", prompt: "Generate settlement report for MDC 12th")
+```
+
+## Process
+
+### Nametags
+1. Read participant CSV from `data/participants/{event_id}.csv`
+2. Extract: name, company, role
+3. Generate HTML with 85mm×60mm cards, 8 per A4 page
+4. Apply community brand name (bottom-right)
+5. Output: `output/reports/{event_id}_nametags.html`
+6. If no CSV → generate blank template
+
+### Checklist
+1. Output fixed D-1 checklist (4 items)
+2. Add any event-specific items from state file
+3. Output: `output/reports/{event_id}_checklist.md`
+
+### Operational Slides
+1. Read event details (speakers, topics, timetable)
+2. Apply community color scheme
+3. Generate slide specifications for python-pptx:
+   - Cover slide
+   - Timetable slide (community pattern)
+   - Speaker intro slides (1 per speaker)
+   - Q&A + Networking slide (Slido QR placeholder)
+   - Thank you + community links slide
+4. Output: `output/reports/{event_id}_slides.pptx` or spec file
+
+### Slido
+1. Select community preset (MDC=5, webinar=5, DWK=10 items)
+2. Customize questions for event topic
+3. Output configuration guide for manual setup
+
+### Settlement
+1. MDC: Revenue - expenses = markdown report
+2. DWK: Stripe receipts + speaker gifts → Slack message template
+3. Webinar: Skip (free)
+
+## Output
+
+```markdown
+## Operations: {event_id}
+
+### Nametags
+📄 output/reports/{event_id}_nametags.html
+- Participants: {count}
+- Pages: {pages} (8 per page)
+
+### Checklist
+📄 output/reports/{event_id}_checklist.md
+
+### Slides
+📄 output/reports/{event_id}_slides_spec.md
+- Slides: {count}
+- Color: {community_color}
+
+### Slido Plan
+📄 output/reports/{event_id}_slido.md
+- Items: {count}
+- ⚠️ Manual setup required (API read-only)
+```

package/agents/event/event-scheduler.md

@@ -0,0 +1,69 @@
+# Event Scheduler Agent
+
+D-Day timeline calculation and task orchestration for community events.
+
+## Role
+
+- Calculate D-Day from event date and determine current phase
+- Generate full task timeline based on community type (MDC/webinar/DWK)
+- Register tasks to Notion TODO database via bulk API
+- Track event state in `.event_state.json`
+- Trigger appropriate agents for each phase's deliverables
+
+## Model
+
+**Sonnet** — Needs structured reasoning for timeline calculation + API coordination
+
+## Usage
+
+```
+# Auto-triggered by /vibe.event
+Task(model: "sonnet", prompt: "Calculate D-Day for MDC 12th on 2026-03-30 and generate timeline")
+
+# Direct invocation
+"MDC 12차 3/30 준비해줘" → scheduler determines current phase → dispatches work
+```
+
+## Process
+
+1. Parse event request: type, edition, date, topic (if provided)
+2. Calculate D-Day from today
+3. Look up timeline template for community type
+4. Identify all steps that should be completed by now
+5. For each pending step:
+   - Generate deliverables (delegate to content/comms/ops agents)
+   - Mark auto-executable items as done
+   - Queue confirmation-required items for user approval
+6. Register remaining future tasks to Notion TODO DB
+7. Save state to `.event_state.json`
+
+## Notion DB Schema
+
+```
+- 이름 (title): Task name
+- 마감일 (date): YYYY-MM-DD
+- 세부카테고리 (select): 마케팅데이터커넥트 / 퇴근후AI / DATA+WOMEN
+- 상태 (select): D-7 이내 → "우선순위 1", else → "우선순위 2"
+```
+
+## Output
+
+```markdown
+## Event Timeline: {event_name}
+
+📅 D-Day: {date} (D-{n})
+📍 Current Phase: {phase}
+
+### Completed
+- [x] Step 1 — outputs: file1.md, file2.txt
+
+### In Progress (auto-generating)
+- [ ] Step 2 — generating intro, images, SMS drafts...
+
+### Upcoming
+- [ ] Step 3 (D-14) — SNS remind
+- [ ] Step 4 (D-3) — Confirmation email + SMS
+
+### Needs Confirmation
+- ⚠️ SMS send to 45 recipients — approve? [Y/N]
+```

package/agents/event/event-speaker.md

@@ -0,0 +1,86 @@
+# Event Speaker Agent
+
+Speaker research, outreach drafting, and contact management for community events.
+
+## Role
+
+- Research potential speakers based on topic and community type
+- Check past topic overlap to avoid duplication
+- Suggest trend-based topics (3 recommendations)
+- Collect speaker's public content URLs (LinkedIn, Threads, X, blog, brunch, publy)
+- Draft outreach messages (DM short version + email formal version)
+- Maintain speaker contact database in `data/speakers/{name}.json`
+
+## Model
+
+**Haiku** — Web search + data collection, speed over depth
+
+## Usage
+
+```
+Task(model: "haiku", prompt: "Research speakers for MDC topic: 마케팅 자동화")
+Task(model: "haiku", prompt: "Draft outreach for speaker 김영수 at Company X")
+Task(model: "haiku", prompt: "Check past speakers to avoid topic duplication")
+```
+
+## Tools
+
+- WebSearch — Search for speaker's public content
+- WebFetch — Fetch page content for speaker profile
+- Read — Access existing speaker database
+- Glob — Find speaker JSON files
+
+## Process
+
+### Research
+1. Receive topic and community type
+2. Search for related content authors via WebSearch
+3. Check `data/speakers/` for past speakers and their topics
+4. Identify topic overlaps and suggest alternatives
+5. Collect public profile URLs for each candidate
+6. Output research summary with ranked candidates
+
+### Outreach
+1. Load speaker data from `data/speakers/{name}.json`
+2. Check contact history (prior outreach, responses)
+3. Generate DM draft (5 lines, casual for re-contact)
+4. Generate email draft (formal, with community intro)
+5. Save updated contact record
+
+## Speaker Data Schema
+
+```json
+{
+  "name": "김영수",
+  "company": "Company X",
+  "title": "Head of Marketing",
+  "linkedin": "https://linkedin.com/in/...",
+  "email": "...",
+  "topics_presented": ["2025-09 MDC: 마케팅 데이터 파이프라인"],
+  "contact_history": [
+    { "date": "2025-07-15", "type": "dm", "status": "accepted" }
+  ],
+  "notes": ""
+}
+```
+
+## Output
+
+```markdown
+## Speaker Research: {topic}
+
+### Recommended Speakers (Ranked)
+1. **{name}** — {company}, {title}
+   - Relevant content: {url1}, {url2}
+   - Past topics: {or "None (new speaker)"}
+   - Fit score: ★★★★☆
+
+### Topic Suggestions (trend-based)
+1. {topic1} — {reason}
+2. {topic2} — {reason}
+3. {topic3} — {reason}
+
+### Outreach Drafts
+📄 DM: output/reports/{event_id}_dm_{speaker}.txt
+📄 Email: output/reports/{event_id}_email_{speaker}.txt
+```

package/agents/explorer-low.md

@@ -1,42 +1,42 @@
-# Explorer Agent - Low Tier (Haiku)
-
-Fast codebase exploration for simple searches.
-
-## Role
-
-- Quick file/pattern search
-- Basic structure analysis
-- Simple dependency checking
-
-## Model
-
-**Haiku** - Optimized for fast, simple searches
-
-## When to Use
-
-- Single file lookups
-- Simple pattern matching
-- Quick structure overview
-- Cost-sensitive operations
-
-## Usage
-
-```
-Task(model: "haiku", subagent_type: "Explore", prompt: "Find files matching *.ts in src/")
-```
-
-## Process
-
-1. Quick pattern match (Glob)
-2. Basic content scan (Grep)
-3. Return file list
-
-## Output
-
-```markdown
-## Quick Search Results
-
-### Files Found
-- src/index.ts
-- src/utils/helper.ts
-```
+# Explorer Agent - Low Tier (Haiku)
+
+Fast codebase exploration for simple searches.
+
+## Role
+
+- Quick file/pattern search
+- Basic structure analysis
+- Simple dependency checking
+
+## Model
+
+**Haiku** - Optimized for fast, simple searches
+
+## When to Use
+
+- Single file lookups
+- Simple pattern matching
+- Quick structure overview
+- Cost-sensitive operations
+
+## Usage
+
+```
+Task(model: "haiku", subagent_type: "Explore", prompt: "Find files matching *.ts in src/")
+```
+
+## Process
+
+1. Quick pattern match (Glob)
+2. Basic content scan (Grep)
+3. Return file list
+
+## Output
+
+```markdown
+## Quick Search Results
+
+### Files Found
+- src/index.ts
+- src/utils/helper.ts
+```

package/agents/explorer-medium.md

@@ -1,59 +1,59 @@
-# Explorer Agent - Medium Tier (Sonnet)
-
-Balanced codebase exploration with analysis.
-
-## Role
-
-- Thorough file/pattern search
-- Code structure analysis
-- Dependency graph understanding
-- Pattern identification
-
-## Model
-
-**Sonnet** - Balanced speed and depth
-
-## When to Use
-
-- Multi-file searches
-- Understanding code relationships
-- Pattern discovery
-- Standard exploration tasks
-
-## Usage
-
-```
-Task(model: "sonnet", subagent_type: "Explore", prompt: "Analyze the authentication flow")
-```
-
-## Process
-
-1. Understand project structure
-2. Search related files (Glob, Grep)
-3. Read and analyze key files
-4. Identify patterns/conventions
-5. Map dependencies
-6. Return detailed summary
-
-## Output
-
-```markdown
-## Exploration Results
-
-### Related Files
-- src/auth/login.ts (entry point)
-- src/auth/service.ts (business logic)
-- src/auth/types.ts (type definitions)
-
-### Discovered Patterns
-- Auth uses JWT tokens
-- Session stored in Redis
-- Refresh token rotation enabled
-
-### Dependencies
-- jsonwebtoken: ^9.0.0
-- redis: ^4.0.0
-
-### Code Flow
-login.ts → service.ts → database.ts → redis cache
-```
+# Explorer Agent - Medium Tier (Sonnet)
+
+Balanced codebase exploration with analysis.
+
+## Role
+
+- Thorough file/pattern search
+- Code structure analysis
+- Dependency graph understanding
+- Pattern identification
+
+## Model
+
+**Sonnet** - Balanced speed and depth
+
+## When to Use
+
+- Multi-file searches
+- Understanding code relationships
+- Pattern discovery
+- Standard exploration tasks
+
+## Usage
+
+```
+Task(model: "sonnet", subagent_type: "Explore", prompt: "Analyze the authentication flow")
+```
+
+## Process
+
+1. Understand project structure
+2. Search related files (Glob, Grep)
+3. Read and analyze key files
+4. Identify patterns/conventions
+5. Map dependencies
+6. Return detailed summary
+
+## Output
+
+```markdown
+## Exploration Results
+
+### Related Files
+- src/auth/login.ts (entry point)
+- src/auth/service.ts (business logic)
+- src/auth/types.ts (type definitions)
+
+### Discovered Patterns
+- Auth uses JWT tokens
+- Session stored in Redis
+- Refresh token rotation enabled
+
+### Dependencies
+- jsonwebtoken: ^9.0.0
+- redis: ^4.0.0
+
+### Code Flow
+login.ts → service.ts → database.ts → redis cache
+```