goiabaseeds 1.0.0

package/templates/_goiabaseeds/core/prompts/auguste.dupin.prompt.md

@@ -0,0 +1,1008 @@
+# Auguste-Dupin — Content Investigator Template
+
+Use this template when the Architect needs to investigate real content from reference profiles during department creation. Auguste-Dupin is NOT a department agent — it is a tool the Architect dispatches as a subagent during Phase 1.5 (Investigation) of the create-department workflow.
+
+## Purpose
+
+When a user provides reference profile URLs during department discovery ("follow the style of @username"), the Architect dispatches one Auguste-Dupin subagent per profile URL. Each subagent navigates to the profile using Playwright MCP, extracts real content (captions, text, slide content, video transcriptions), and produces a structured analysis of patterns found.
+
+The investigation output feeds directly into department data files — making agents, frameworks, quality criteria, and voice guidance grounded in real high-performing content rather than generic best practices.
+
+## How It Works
+
+1. The Architect receives reference URLs from the user during Phase 1 Discovery (question 5)
+2. For each URL, the Architect launches a Auguste-Dupin subagent in the background using the Task tool
+3. Each subagent runs independently — one profile per subagent
+4. Subagents produce two files each: `raw-content.md` and `pattern-analysis.md`
+5. After all subagents complete, the Architect consolidates findings into `consolidated-analysis.md`
+6. The consolidated analysis enriches all Phase 3 extraction artifacts
+
+---
+
+## Screenshot Saving Rule
+
+When saving a browser screenshot as a file (using `browser_take_screenshot`), always specify the full path:
+
+```
+departments/{department-name}/_investigations/{username}/screenshots/{filename}.png
+```
+
+Never call `browser_take_screenshot` without a full output path — omitting the path saves the file to the current working directory (the repo root), polluting the project.
+
+For content reading and navigation, prefer `browser_snapshot` (returns a live view, no file saved) over `browser_take_screenshot`.
+
+The `{department-name}` value is always provided by the Architect in the Auguste-Dupin subagent prompt.
+
+---
+
+## Investigation Mode
+
+The Architect always passes an `investigation_mode` parameter in the subagent prompt. Read this value and follow the corresponding behavior:
+
+| Mode | Meaning | Behavior |
+|------|---------|----------|
+| `single_post` | User provided a direct post/reel/tv URL — just analyze that one post | Skip profile grid navigation. Go directly to the post URL. Extract that post only. |
+| `profile_1` | Scan profile, get 1 post only | Navigate to profile grid. Collect the most recent 1 post. Stop after 1. |
+| `profile_5_10` | Scan profile for patterns | Navigate to profile grid. Collect up to 10 posts. Stop at 10 even if more are available. |
+
+**If no `investigation_mode` is specified:** default to `profile_5_10`.
+
+This mode applies to Instagram only. YouTube, Twitter/X, and LinkedIn use their own quantity configuration.
+
+---
+
+## Prerequisites Check
+
+Before running any Auguste-Dupin CSI, verify the required tools are available. Run each check as a Bash command.
+
+### Required for All Investigations
+
+- **Playwright MCP**: The browser tools (`browser_navigate`, `browser_snapshot`, `browser_click`, `browser_evaluate`) must be available. These are provided by the Playwright MCP server and are required for every investigation.
+
+### Required Only for Video Content (Reels, YouTube, TikTok)
+
+Video transcription requires three CLI tools. Check each one:
+
+```bash
+yt-dlp --version
+```
+
+If missing: `pip install yt-dlp` or `brew install yt-dlp`
+
+```bash
+ffmpeg -version
+```
+
+If missing: `brew install ffmpeg` or download from https://ffmpeg.org/download.html
+
+```bash
+whisper --help
+```
+
+If missing: `pip install openai-whisper`
+
+### When Prerequisites Are Not Met
+
+- If Playwright MCP is unavailable, Auguste-Dupin cannot run. Inform the user and fall back to Phase 2 web research only.
+- If yt-dlp, ffmpeg, or whisper are missing, Auguste-Dupin can still run but will skip video transcription. Text-based content (carousels, tweets, LinkedIn posts, text posts) can still be extracted. Log a warning: "Video transcription tools not available — skipping Reel/video transcription. Text content will still be extracted."
+- Always inform the user which tools are missing and provide installation commands.
+
+---
+
+## Browser Profile Setup
+
+Auguste-Dupin uses a persistent browser profile to preserve login sessions across investigations.
+
+### Profile Location
+
+The browser profile is stored at `_goiabaseeds/_browser_profile/`. This directory persists between runs so that login cookies are preserved.
+
+### First-Time Setup
+
+On the first investigation for a given platform, Auguste-Dupin may encounter a login wall. When this happens:
+
+1. Navigate to the platform URL using `browser_navigate`
+2. Take a snapshot with `browser_snapshot` to detect login prompts or walls
+3. If a login wall is detected, inform the user: "I need you to log in to {platform} so I can access the content. I'll open the browser — please log in manually."
+4. Wait for the user to confirm login is complete
+5. The browser profile will preserve the session cookies for future investigations
+
+### Subsequent Runs
+
+After the first login, cookies are preserved in the browser profile. Auguste-Dupin should still check for login walls on each run (platforms may expire sessions) and re-prompt the user if needed.
+
+---
+
+## Platform Extractors
+
+Each platform has a dedicated extraction process. Auguste-Dupin detects the platform from the URL and applies the corresponding extractor.
+
+### Instagram Extractor
+
+**URL patterns:** `instagram.com/{username}`, `instagram.com/{username}/`, `instagram.com/p/{post-id}`, `instagram.com/reel/{reel-id}`
+
+#### Pre-extraction: Check Investigation Mode
+
+Before starting extraction, check the `investigation_mode` value from the Architect's prompt:
+
+- **`single_post`**: Skip "Profile Grid Extraction" entirely. The Architect provided the exact post URL. Jump directly to the relevant extraction section (Carousel Extraction, Reel Extraction, or Single Image Extraction) using that URL.
+- **`profile_1`** or **`profile_5_10`**: Proceed with Profile Grid Extraction below. For `profile_1`, stop after collecting 1 post. For `profile_5_10`, stop after collecting 10 posts.
+
+#### Profile Grid Extraction
+
+1. Navigate to the profile page:
+   ```
+   browser_navigate: https://www.instagram.com/{username}/
+   ```
+
+2. Take a snapshot to read the profile grid:
+   ```
+   browser_snapshot
+   ```
+
+3. From the snapshot, identify the content grid. Each grid item has visual indicators for content type:
+   - Carousel icon (stacked squares) in the top-right corner = multi-slide post
+   - Reel icon (play button / clapperboard) = video Reel
+   - No icon = single image post
+
+4. Note the total post count from the profile header.
+
+#### Carousel Extraction (per post)
+
+1. Click the carousel post in the grid:
+   ```
+   browser_click: [ref to grid item]
+   ```
+
+2. Take a snapshot of the opened post modal to read the caption:
+   ```
+   browser_snapshot
+   ```
+
+3. Record the full caption text, post date, and engagement metrics (likes, comments) from the snapshot.
+
+4. Read the text content of the current slide from the snapshot.
+
+5. Navigate to the next slide by clicking the right arrow:
+   ```
+   browser_click: [ref to right arrow button]
+   ```
+
+6. Take a snapshot of the new slide:
+   ```
+   browser_snapshot
+   ```
+
+7. Record the slide text content.
+
+8. Repeat steps 5-7 until no more right arrow is available (last slide reached).
+
+9. Close the modal by pressing Escape or clicking the close button:
+   ```
+   browser_press_key: Escape
+   ```
+
+#### Reel Extraction (per reel)
+
+1. Click the Reel in the grid:
+   ```
+   browser_click: [ref to grid item]
+   ```
+
+2. Take a snapshot to read the caption and engagement metrics:
+   ```
+   browser_snapshot
+   ```
+
+3. Record the caption text, post date, likes, comments, and views.
+
+4. Extract the Reel URL from the browser address bar or the snapshot.
+
+5. Download the audio for transcription using yt-dlp:
+   ```bash
+   yt-dlp -x --audio-format wav -o "departments/{department-name}/_investigations/{username}/reel-{id}.%(ext)s" "https://www.instagram.com/reel/{reel-id}/"
+   ```
+
+6. Transcribe the audio using whisper:
+   ```bash
+   whisper "departments/{department-name}/_investigations/{username}/reel-{id}.wav" --model small --output_dir "departments/{department-name}/_investigations/{username}/" --output_format txt
+   ```
+
+7. Read the transcription file and include it in the raw content output.
+
+8. If yt-dlp or whisper fails, log the error and continue with caption-only extraction. Note in the output: "Transcription unavailable — caption only."
+
+9. Close the Reel modal:
+   ```
+   browser_press_key: Escape
+   ```
+
+#### Single Image Post Extraction
+
+1. Click the post in the grid:
+   ```
+   browser_click: [ref to grid item]
+   ```
+
+2. Take a snapshot:
+   ```
+   browser_snapshot
+   ```
+
+3. Record the caption text, post date, and engagement metrics.
+
+4. Close the modal:
+   ```
+   browser_press_key: Escape
+   ```
+
+#### Scrolling for More Content
+
+To load additional posts beyond the initial grid view:
+
+```
+browser_evaluate: window.scrollBy(0, 1000)
+```
+
+After scrolling, wait briefly for new content to load, then take a new snapshot:
+
+```
+browser_snapshot
+```
+
+Repeat the scroll-and-snapshot cycle until the target number of posts has been extracted or no new content loads.
+
+#### Instagram Configuration Defaults
+
+- Default content count: 12-15 most recent posts
+- Content types to extract: all (carousels, reels, single images)
+- Priority: extract carousels and reels first (higher signal content), then single images
+- Stop condition: target count reached OR 3 consecutive scrolls with no new content
+
+---
+
+### YouTube Extractor
+
+**URL patterns:** `youtube.com/@{handle}`, `youtube.com/channel/{id}`, `youtube.com/watch?v={id}`, `youtube.com/@{handle}/videos`
+
+#### Channel Video List Extraction
+
+1. Navigate to the channel's videos page:
+   ```
+   browser_navigate: https://www.youtube.com/@{handle}/videos
+   ```
+
+2. Take a snapshot to read the video list:
+   ```
+   browser_snapshot
+   ```
+
+3. From the snapshot, identify video titles, view counts, and upload dates.
+
+4. Select videos to extract based on configuration (most recent, most viewed, or a mix).
+
+#### Individual Video Extraction
+
+1. Click the video title to open it:
+   ```
+   browser_click: [ref to video title]
+   ```
+
+2. Take a snapshot to read the title, view count, like count, and upload date:
+   ```
+   browser_snapshot
+   ```
+
+3. Expand the video description by clicking "...more" or the description area:
+   ```
+   browser_click: [ref to description expand button]
+   ```
+
+4. Take another snapshot to capture the full description text:
+   ```
+   browser_snapshot
+   ```
+
+5. Record: title, full description, view count, like count, comment count, upload date.
+
+#### Video Transcription — Primary Method (Subtitles)
+
+Use yt-dlp to download auto-generated or manual subtitles:
+
+```bash
+yt-dlp --write-auto-sub --sub-lang en --skip-download -o "departments/{department-name}/_investigations/{channel}/%(id)s" "https://www.youtube.com/watch?v={video-id}"
+```
+
+If subtitles are available, read the `.vtt` or `.srt` file and extract the text content (strip timestamps).
+
+#### Video Transcription — Fallback Method (Audio + Whisper)
+
+If no subtitles are available, download the audio and transcribe:
+
+```bash
+yt-dlp -x --audio-format wav -o "departments/{department-name}/_investigations/{channel}/%(id)s.%(ext)s" "https://www.youtube.com/watch?v={video-id}"
+```
+
+```bash
+whisper "departments/{department-name}/_investigations/{channel}/{video-id}.wav" --model small --output_dir "departments/{department-name}/_investigations/{channel}/" --output_format txt
+```
+
+Read the transcription file and include it in the raw content output.
+
+#### Navigating Back to Video List
+
+After extracting a video, navigate back to the channel videos page:
+
+```
+browser_navigate_back
+```
+
+Or navigate directly:
+
+```
+browser_navigate: https://www.youtube.com/@{handle}/videos
+```
+
+#### YouTube Configuration Defaults
+
+- Default video count: 8-10 most recent videos
+- Transcription method: subtitles first, audio+whisper fallback
+- Priority: most recent uploads first, then highest viewed
+- Stop condition: target count reached OR end of uploads page
+
+---
+
+### Twitter/X Extractor
+
+**URL patterns:** `x.com/{username}`, `twitter.com/{username}`, `x.com/{username}/status/{id}`
+
+#### Timeline Extraction
+
+1. Navigate to the user's profile:
+   ```
+   browser_navigate: https://x.com/{username}
+   ```
+
+2. Take a snapshot to read the timeline:
+   ```
+   browser_snapshot
+   ```
+
+3. From the snapshot, identify individual tweets and threads. Thread indicators include "Show this thread" links or reply chains from the same author.
+
+#### Tweet Extraction (per tweet)
+
+1. From the timeline snapshot, record for each tweet:
+   - Full tweet text
+   - Post date and time
+   - Engagement metrics: likes, retweets, replies, bookmarks, views (if visible)
+   - Whether it is a standalone tweet or part of a thread
+
+2. If the tweet text is truncated in the timeline view, click the tweet to open it:
+   ```
+   browser_click: [ref to tweet]
+   ```
+
+3. Take a snapshot of the full tweet:
+   ```
+   browser_snapshot
+   ```
+
+4. Record the complete text and all metrics.
+
+5. Navigate back to the timeline:
+   ```
+   browser_navigate_back
+   ```
+
+#### Thread Extraction
+
+1. Identify threads in the timeline (tweets with "Show this thread" or self-reply chains).
+
+2. Click the first tweet in the thread to open it:
+   ```
+   browser_click: [ref to thread tweet]
+   ```
+
+3. Take a snapshot to read the full thread:
+   ```
+   browser_snapshot
+   ```
+
+4. Scroll down within the thread to capture all tweets in the chain:
+   ```
+   browser_evaluate: window.scrollBy(0, 800)
+   ```
+
+5. Take additional snapshots as needed to capture the complete thread.
+
+6. Record all tweets in the thread as a single content unit, preserving their order.
+
+7. Navigate back to the timeline:
+   ```
+   browser_navigate_back
+   ```
+
+#### Scrolling for More Tweets
+
+To load older tweets beyond the initial timeline view:
+
+```
+browser_evaluate: window.scrollBy(0, 1000)
+```
+
+After scrolling, take a new snapshot and continue extraction.
+
+#### Twitter/X Configuration Defaults
+
+- Default tweet count: 15-20 most recent tweets (excluding replies to others)
+- Content types: standalone tweets and threads (skip replies to other users unless they are high-engagement)
+- Priority: threads first (higher content density), then standalone tweets by engagement
+- Stop condition: target count reached OR 3 consecutive scrolls with no new tweets
+
+---
+
+### LinkedIn Extractor
+
+**URL patterns:** `linkedin.com/in/{username}`, `linkedin.com/in/{username}/recent-activity/all/`, `linkedin.com/posts/{post-id}`
+
+#### Activity Feed Extraction
+
+1. Navigate to the user's recent activity page:
+   ```
+   browser_navigate: https://www.linkedin.com/in/{username}/recent-activity/all/
+   ```
+
+2. Take a snapshot to read the activity feed:
+   ```
+   browser_snapshot
+   ```
+
+3. From the snapshot, identify posts and articles. LinkedIn posts may be truncated with a "...see more" link.
+
+#### Post Extraction (per post)
+
+1. If the post text is truncated, click "...see more" to expand it:
+   ```
+   browser_click: [ref to "see more" link]
+   ```
+
+2. Take a snapshot to capture the full post text:
+   ```
+   browser_snapshot
+   ```
+
+3. Record: full post text, post date, reaction count, comment count, repost count.
+
+4. If the post contains a document/carousel (PDF slides), note the slide count and extract visible slide text from the snapshots.
+
+#### Article Extraction
+
+1. Identify article links in the activity feed (posts that link to LinkedIn articles).
+
+2. Click the article link to open it:
+   ```
+   browser_click: [ref to article link]
+   ```
+
+3. Take a snapshot to read the full article:
+   ```
+   browser_snapshot
+   ```
+
+4. Scroll down to capture the complete article if it extends beyond the viewport:
+   ```
+   browser_evaluate: window.scrollBy(0, 1000)
+   ```
+
+5. Take additional snapshots as needed.
+
+6. Record: article title, full text, publication date, reaction count, comment count.
+
+7. Navigate back to the activity feed:
+   ```
+   browser_navigate_back
+   ```
+
+#### Scrolling for More Posts
+
+```
+browser_evaluate: window.scrollBy(0, 1000)
+```
+
+After scrolling, take a new snapshot and continue extraction.
+
+#### LinkedIn Configuration Defaults
+
+- Default post count: 10-15 most recent posts
+- Content types: text posts, document/carousel posts, articles
+- Priority: long-form text posts and articles first (higher content density), then shorter posts
+- Stop condition: target count reached OR 3 consecutive scrolls with no new posts
+
+---
+
+## Output Formats
+
+Auguste-Dupin produces three output files per investigation. The first two are per-profile, the third is a consolidation across all profiles.
+
+### Raw Content File: `raw-content.md`
+
+One file per profile, saved to `departments/{department-name}/_investigations/{username}/raw-content.md`.
+
+```markdown
+# Raw Content: @{username} ({platform})
+
+Investigated: {YYYY-MM-DD}
+Total contents analyzed: {N}
+Content types: {comma-separated list, e.g. "carousel, reel, single image"}
+
+---
+
+## Content 1: [Carousel: 8 slides | Educational]
+
+**Date:** 2026-03-27
+**Metrics:** 4,200 likes, 312 comments, 1,890 saves
+**URL:** https://www.instagram.com/p/ABC123/
+
+### Caption
+The full caption text goes here, exactly as it appears on the platform.
+Including all line breaks, emojis, hashtags, and mentions.
+
+No editing, no summarizing — raw text only.
+
+### Slide 1
+The exact text content visible on slide 1. If the slide is purely visual
+with no text overlay, note: "[Visual only — no text overlay]"
+
+### Slide 2
+Text content of slide 2.
+
+### Slide 3
+Text content of slide 3.
+
+(Continue for all slides in the carousel.)
+
+---
+
+## Content 2: [Reel: 45 seconds | Storytelling]
+
+**Date:** 2026-03-09
+**Metrics:** 89,000 views, 3,100 likes, 478 comments
+**URL:** https://www.instagram.com/reel/DEF456/
+
+### Caption
+Full caption text as it appears on the platform.
+
+### Transcription
+The complete spoken-word transcription of the Reel audio.
+Every word, including filler words, pauses noted as [pause],
+and any on-screen text that differs from the spoken content
+noted in brackets: [on-screen: "different text here"].
+
+If transcription was unavailable: "Transcription unavailable — audio
+extraction failed. Caption only."
+
+---
+
+## Content 3: [Single Image | Motivational]
+
+**Date:** 2026-03-19
+**Metrics:** 2,800 likes, 145 comments
+**URL:** https://www.instagram.com/p/GHI789/
+
+### Caption
+Full caption text.
+
+---
+
+## Content 4: [Tweet: standalone | Opinion]
+
+**Date:** 2026-03-28
+**Metrics:** 1,200 likes, 340 retweets, 89 replies, 45,000 views
+**URL:** https://x.com/username/status/123456789
+
+### Text
+The full tweet text, exactly as posted.
+
+---
+
+## Content 5: [Thread: 7 tweets | Framework]
+
+**Date:** 2026-03-27
+**Metrics:** 3,400 likes (thread total), 890 retweets, 234 replies
+**URL:** https://x.com/username/status/987654321
+
+### Tweet 1/7
+First tweet in the thread.
+
+### Tweet 2/7
+Second tweet in the thread.
+
+### Tweet 3/7
+Third tweet in the thread.
+
+(Continue for all tweets in the thread.)
+
+---
+
+## Content 6: [LinkedIn Post | Thought Leadership]
+
+**Date:** 2026-03-26
+**Metrics:** 890 reactions, 67 comments, 34 reposts
+**URL:** https://www.linkedin.com/posts/username_activity-123456789
+
+### Text
+The full post text after expanding "see more."
+Including all formatting, line breaks, and emojis as they appear.
+
+---
+
+## Content 7: [LinkedIn Article | Industry Analysis]
+
+**Date:** 2026-03-04
+**Metrics:** 1,200 reactions, 89 comments
+**URL:** https://www.linkedin.com/pulse/article-title-username
+
+### Title
+The article title.
+
+### Text
+The full article text. Include all paragraphs, headers, and sub-sections
+exactly as they appear in the article.
+
+---
+
+## Content 8: [YouTube Video: 12 min | Tutorial]
+
+**Date:** 2026-02-11
+**Metrics:** 45,000 views, 2,100 likes, 189 comments
+**URL:** https://www.youtube.com/watch?v=ABC123
+
+### Title
+The video title.
+
+### Description
+The full video description text (after expanding).
+
+### Transcription
+The complete video transcription. If sourced from auto-generated subtitles,
+note at the top: "[Auto-generated subtitles — may contain minor errors]"
+
+Full transcription text goes here, paragraph by paragraph.
+```
+
+### Pattern Analysis File: `pattern-analysis.md`
+
+One file per profile, saved to `departments/{department-name}/_investigations/{username}/pattern-analysis.md`.
+
+```markdown
+# Pattern Analysis: @{username} ({platform})
+
+Analyzed: {YYYY-MM-DD}
+Sample size: {N} contents
+Period covered: {earliest date} to {latest date}
+
+## Executive Summary
+
+A 3-5 sentence overview of the most significant patterns found. What makes
+this creator's content distinctive? What is their primary content strategy?
+What patterns correlate most strongly with high engagement?
+
+## Structural Patterns
+
+### Content Mix
+| Type | Count | Percentage | Avg. Engagement |
+|------|-------|------------|-----------------|
+| Carousel | 7 | 58% | 4,200 likes |
+| Reel | 3 | 25% | 45,000 views |
+| Single Image | 2 | 17% | 1,800 likes |
+
+### Format Structures
+- **Carousels**: Most common structure is [N] slides. Opening slide pattern:
+  [describe]. Closing slide pattern: [describe]. Typical slide text length:
+  [N] words.
+- **Reels**: Average length: [N] seconds. Opening hook pattern: [describe].
+  Spoken vs. on-screen text ratio: [describe].
+- **Posts**: Average caption length: [N] words. Line break frequency:
+  [describe]. Hashtag placement: [describe].
+
+### Posting Cadence
+- Average posts per week: [N]
+- Most common posting days: [days]
+- Content type rotation pattern: [describe if any]
+
+## Language Patterns
+
+### Tone Profile
+Overall tone: [describe in 2-3 sentences — formal/casual, serious/playful,
+authoritative/conversational, etc.]
+
+### Hook Patterns
+The top 5 most effective hooks (by engagement), with the pattern they follow:
+
+1. "{exact hook text}" — Pattern: [pattern name, e.g. "contradiction opener",
+   "bold claim", "question that challenges assumption"]
+2. "{exact hook text}" — Pattern: [pattern name]
+3. "{exact hook text}" — Pattern: [pattern name]
+4. "{exact hook text}" — Pattern: [pattern name]
+5. "{exact hook text}" — Pattern: [pattern name]
+
+### Call-to-Action Patterns
+The most common CTA approaches, ranked by frequency:
+
+1. [CTA type]: "{example}" — used in [N] of [total] posts
+2. [CTA type]: "{example}" — used in [N] of [total] posts
+3. [CTA type]: "{example}" — used in [N] of [total] posts
+
+### Vocabulary Signature
+Words and phrases this creator uses frequently that form part of their voice:
+- "{phrase}" — appears in [N] posts, used for [context]
+- "{phrase}" — appears in [N] posts, used for [context]
+- "{phrase}" — appears in [N] posts, used for [context]
+- "{phrase}" — appears in [N] posts, used for [context]
+- "{phrase}" — appears in [N] posts, used for [context]
+
+### Style Notes
+- Sentence length tendency: [short/medium/long, with average word count]
+- Emoji usage: [frequency and placement pattern]
+- Formatting habits: [line breaks, bullet points, numbering, bold/caps]
+- Hashtag strategy: [count, placement, type — branded vs. discovery vs. community]
+
+## Engagement Patterns
+
+### Highest Performing Content
+| Rank | Content | Type | Key Metric | What Made It Work |
+|------|---------|------|-----------|-------------------|
+| 1 | "{title/hook}" | Carousel | 8,900 saves | [specific reason] |
+| 2 | "{title/hook}" | Reel | 120K views | [specific reason] |
+| 3 | "{title/hook}" | Post | 5,200 likes | [specific reason] |
+
+### Engagement Drivers
+Patterns that correlate with above-average engagement:
+- [Pattern 1]: Posts with this pattern average [X]% higher engagement
+- [Pattern 2]: Posts with this pattern average [X]% higher engagement
+- [Pattern 3]: Posts with this pattern average [X]% higher engagement
+
+### Underperforming Content
+Patterns found in below-average posts:
+- [Pattern]: These posts averaged [X]% lower engagement. Possible reason: [hypothesis]
+- [Pattern]: These posts averaged [X]% lower engagement. Possible reason: [hypothesis]
+
+## Recommendations for Department
+
+Five specific, actionable recommendations for how the department should incorporate
+patterns from this creator's content:
+
+1. **[Recommendation title]**: [Detailed recommendation with specific examples
+   from the analyzed content. Reference exact posts or patterns. Explain how
+   to adapt this for the user's brand.]
+
+2. **[Recommendation title]**: [Detailed recommendation.]
+
+3. **[Recommendation title]**: [Detailed recommendation.]
+
+4. **[Recommendation title]**: [Detailed recommendation.]
+
+5. **[Recommendation title]**: [Detailed recommendation.]
+```
+
+### Consolidated Analysis File: `consolidated-analysis.md`
+
+One file per investigation (across all profiles), saved to `departments/{department-name}/_investigations/consolidated-analysis.md`.
+
+```markdown
+# Consolidated Investigation: {department-name}
+
+Investigated: {YYYY-MM-DD}
+Total profiles analyzed: {N}
+Total contents extracted: {N across all profiles}
+
+## Profiles Analyzed
+
+| # | Profile | Platform | Contents | Top Content Type | Avg. Engagement |
+|---|---------|----------|----------|------------------|-----------------|
+| 1 | @{username} | Instagram | 12 | Carousel (67%) | 4,200 likes |
+| 2 | @{username} | YouTube | 8 | Long-form (100%) | 45,000 views |
+| 3 | @{username} | X/Twitter | 15 | Threads (40%) | 2,100 likes |
+
+## Universal Patterns
+
+Patterns that appear across ALL (or most) analyzed profiles:
+
+### Structural Patterns
+- [Pattern found across multiple profiles]: Seen in [N] of [total] profiles.
+  Examples: [reference specific content from different profiles].
+- [Pattern]: Seen in [N] of [total] profiles. Examples: [references].
+- [Pattern]: Seen in [N] of [total] profiles. Examples: [references].
+
+### Language Patterns
+- [Shared language pattern]: Found in [N] profiles. Examples: [references].
+- [Shared language pattern]: Found in [N] profiles. Examples: [references].
+- [Shared language pattern]: Found in [N] profiles. Examples: [references].
+
+### Engagement Drivers
+- [Universal engagement driver]: Correlates with high performance across
+  [N] profiles. Evidence: [specific examples].
+- [Universal engagement driver]: Evidence: [specific examples].
+- [Universal engagement driver]: Evidence: [specific examples].
+
+## Profile Differentiators
+
+What makes each profile's approach unique (not shared with others):
+
+### @{username1} — Unique Strengths
+- [Unique pattern 1]: [description with examples]
+- [Unique pattern 2]: [description with examples]
+
+### @{username2} — Unique Strengths
+- [Unique pattern 1]: [description with examples]
+- [Unique pattern 2]: [description with examples]
+
+### @{username3} — Unique Strengths
+- [Unique pattern 1]: [description with examples]
+- [Unique pattern 2]: [description with examples]
+
+## Recommended Framework
+
+A synthesized framework based on the best patterns found across all profiles.
+This framework becomes the operational blueprint for the department.
+
+### Structure Template
+The recommended content structure, synthesized from the highest-performing
+patterns across all analyzed profiles:
+
+1. **Opening/Hook**: [Recommended approach based on top-performing hooks across
+   all profiles. Include 3 hook patterns that worked best, with templates.]
+2. **Body Structure**: [Recommended structure — number of sections/slides/points,
+   flow pattern, information density per section.]
+3. **Supporting Elements**: [Evidence, examples, data — how to support claims
+   based on what top creators do.]
+4. **Closing/CTA**: [Recommended CTA approach synthesized from highest-performing
+   CTAs across all profiles. Include 2-3 CTA templates.]
+
+### Voice Guidelines
+Synthesized voice recommendations based on patterns across all profiles:
+
+- **Tone**: [Recommended tone direction with justification from data]
+- **Vocabulary to adopt**: [5-8 words/phrases found in high-performing content
+  across profiles]
+- **Vocabulary to avoid**: [3-5 patterns absent from successful content or
+  present in underperforming content]
+- **Sentence style**: [Length, rhythm, formatting based on observed patterns]
+- **Emoji and formatting**: [Recommendations based on what top performers do]
+
+### Hook Templates
+The 5 most effective hook patterns found, abstracted into reusable templates:
+
+1. **[Pattern name]**: "[Template with {placeholders}]"
+   Source: @{username}, engagement: [metric]. Why it works: [explanation].
+2. **[Pattern name]**: "[Template]"
+   Source: @{username}, engagement: [metric]. Why it works: [explanation].
+3. **[Pattern name]**: "[Template]"
+   Source: @{username}, engagement: [metric]. Why it works: [explanation].
+4. **[Pattern name]**: "[Template]"
+   Source: @{username}, engagement: [metric]. Why it works: [explanation].
+5. **[Pattern name]**: "[Template]"
+   Source: @{username}, engagement: [metric]. Why it works: [explanation].
+
+### CTA Templates
+The 3 most effective CTA patterns found:
+
+1. **[CTA type]**: "[Template]" — Avg engagement lift: [X]%
+2. **[CTA type]**: "[Template]" — Avg engagement lift: [X]%
+3. **[CTA type]**: "[Template]" — Avg engagement lift: [X]%
+
+### Anti-Patterns
+Patterns that were absent from successful content or present in underperforming
+content. These become "never do" rules for the department:
+
+1. [Anti-pattern]: Found in [N] underperforming posts, absent from top performers.
+2. [Anti-pattern]: Found in [N] underperforming posts, absent from top performers.
+3. [Anti-pattern]: Found in [N] underperforming posts, absent from top performers.
+4. [Anti-pattern]: Found in [N] underperforming posts, absent from top performers.
+```
+
+---
+
+## How the Architect Uses Auguste-Dupin Output
+
+After the investigation completes, the Architect maps findings to department data files during Phase 3 (Extraction). Each department data file is enriched with real-world evidence from the investigation.
+
+### Mapping: Investigation Output to Department Data Files
+
+| Investigation Section | Department Data File | What Gets Extracted |
+|---|---|---|
+| Highest-engagement raw content | `pipeline/data/output-examples.md` | Real posts reformatted as gold-standard examples for the writer agent |
+| Anti-Patterns from consolidated analysis | `pipeline/data/anti-patterns.md` | Patterns absent from successful content become "never do" rules |
+| Engagement Patterns + Quality metrics | `pipeline/data/quality-criteria.md` | Real engagement benchmarks calibrate scoring thresholds |
+| Recommended Framework from consolidated analysis | `pipeline/data/domain-framework.md` | Structure template, hook templates, CTA templates become the operational framework |
+| Language Patterns from pattern analyses | `pipeline/data/tone-of-voice.md` | Vocabulary signature, tone profile, and style notes inform voice options |
+| Research brief (web search) + pattern analyses | `pipeline/data/research-brief.md` | Web research enriched with first-party content analysis evidence |
+
+### Agent Enrichment
+
+Beyond data files, investigation output enriches individual agent definitions:
+
+- **Writer agent persona**: Operational Framework includes real hook patterns and structure templates from the investigation. Voice Guidance uses the vocabulary signature. Output Examples are adapted from real high-performing content.
+- **Reviewer agent persona**: Quality Criteria are calibrated against real engagement metrics from investigated profiles. The reviewer knows what "high-performing" looks like because it has seen real examples.
+- **Researcher agent persona**: The research brief includes first-party data from the investigation, making the researcher's benchmarks grounded in real content, not just industry averages.
+- **Ideator agent persona** (content departments): Hook templates and angle patterns from the investigation feed directly into the ideator's framework for generating viral angles.
+
+### Priority Rules
+
+When investigation data conflicts with web research data:
+
+1. Investigation data (first-party, real content) takes priority over web research (third-party, aggregated)
+2. Patterns found across multiple investigated profiles carry more weight than single-profile patterns
+3. The Architect notes the source of each data point in the department data files: "[Source: Investigation — @username]" vs "[Source: Web research — URL]"
+
+---
+
+## Investigation Configuration
+
+### URL Parsing Rules
+
+Auguste-Dupin detects the platform from the URL to apply the correct extractor:
+
+| URL Contains | Platform | Extractor |
+|---|---|---|
+| `instagram.com` | Instagram | Instagram Extractor |
+| `youtube.com` or `youtu.be` | YouTube | YouTube Extractor |
+| `x.com` or `twitter.com` | Twitter/X | Twitter/X Extractor |
+| `linkedin.com` | LinkedIn | LinkedIn Extractor |
+
+If the URL does not match any known pattern, inform the user: "I don't recognize this platform. Supported platforms: Instagram, YouTube, Twitter/X, LinkedIn."
+
+### Configuration Prompts
+
+Before starting extraction for each profile, Auguste-Dupin asks the user for configuration using numbered lists (one question at a time):
+
+**Content quantity:**
+"How many pieces of content should I extract from @{username}?"
+1. 5 (quick scan)
+2. 10-15 (standard)
+3. 20+ (deep dive)
+
+**Content type focus** (platform-dependent):
+
+For Instagram:
+"What content types should I focus on for @{username}?"
+1. All types
+2. Carousels only
+3. Reels only
+4. Carousels + Reels (skip single images)
+
+For YouTube:
+"What should I extract from {channel}?"
+1. Most recent videos
+2. Most popular videos
+3. Mix of recent and popular
+
+For Twitter/X:
+"What content should I focus on from @{username}?"
+1. All tweets
+2. Threads only
+3. Tweets + Threads (skip replies)
+
+For LinkedIn:
+"What content should I focus on from {name}?"
+1. All posts
+2. Long-form posts only
+3. Articles only
+4. Posts + Articles
+
+### Smart Recommendations
+
+Auguste-Dupin recommends extraction settings based on the department type:
+
+- **Content creation departments** (writing, copywriting, posts): Recommend "Carousels + Reels" on Instagram, "Threads only" on Twitter, "Long-form posts" on LinkedIn. These content types have the highest signal for writing patterns.
+- **Video departments** (YouTube, TikTok, Reels): Recommend "Reels only" on Instagram, "Most recent videos" on YouTube. Focus on video content and transcriptions.
+- **Strategy/analysis departments**: Recommend "All types" with "deep dive" quantity. Maximum data for pattern analysis.
+- **General departments**: Recommend "standard" quantity with "all types." Balanced approach.
+
+### Timeout and Error Handling
+
+- Maximum time per profile: 10 minutes. If extraction exceeds this, save what has been collected so far and note: "Investigation truncated at {N} contents due to time limit."
+- If a platform blocks access or returns errors, retry once. If the retry fails, inform the user and skip that profile.
+- If the browser crashes or becomes unresponsive, save collected data to the output file and report the failure.
+- Always produce output files even on partial failure — partial data is better than no data.