vibesuite 1.0.0

package/.agent/skills/youtube-pipeline/resources/youtube-phase5-repurposing.md

@@ -0,0 +1,159 @@
+---
+description: Phase 5 of YouTube Pipeline - Repurposing & Distribution. Cross-post content to all platforms with strategic scheduling.
+---
+
+# YouTube Phase 5: Repurposing & Distribution
+
+> **Your Role**: You are the Distribution Strategist. You ensure every piece of content reaches every platform, maximizing reach from a single filming session.
+
+## When to Run This
+
+Run this **after Phase 4 (Production)** when the main video and shorts are ready for upload.
+
+---
+
+## The 2-Week Content Calendar
+
+After filming one main video, you have **7 pieces of content** to distribute:
+
+| Day | YouTube | TikTok | Instagram | LinkedIn |
+|-----|---------|--------|-----------|----------|
+| **Monday (Day 1)** | Main Video + Topic Short 1 | Topic Short 1 | Topic Short 1 (Reel) | Topic Short 1 + Text Post 1 |
+| **Wednesday** | Topic Short 2 | Topic Short 2 | Topic Short 2 (Reel) | Topic Short 2 |
+| **Friday** | Topic Short 3 | Topic Short 3 | Topic Short 3 (Reel) | Topic Short 3 + Text Post 2 |
+| **Next Monday** | Extract Short 1 | Extract Short 1 | Extract Short 1 (Reel) | Extract Short 1 |
+| **Next Wednesday** | Extract Short 2 | Extract Short 2 | Extract Short 2 (Reel) | Extract Short 2 + Text Post 3 |
+| **Next Friday** | Extract Short 3 | Extract Short 3 | Extract Short 3 (Reel) | Extract Short 3 |
+
+---
+
+## Platform-Specific Notes
+
+### YouTube
+- Main video: Full description, tags, end screen, cards
+- Shorts: Post to YouTube Shorts naturally (under 60 seconds, vertical)
+
+### TikTok & Instagram Reels
+- **Same content as YouTube Shorts** — just repost
+- Remove any YouTube-specific CTAs (like "link in description")
+- Use platform-native captions/hashtags
+
+### LinkedIn (Multi-Format)
+
+LinkedIn gets the **most content variety**:
+
+| Content Type | Quantity | Schedule |
+|--------------|----------|----------|
+| **Video Shorts** | All 6 | Same schedule as YouTube |
+| **Text Posts** | 3 | Mon/Fri/Wed of Week 1 and Week 2 |
+| **Stories** | As available | Behind-the-scenes, teasers |
+
+#### LinkedIn Text Post Templates
+
+**Post 1: The Hook/Pain (Monday with main video)**
+```
+[Attention-grabbing first line about the problem]
+
+Here's what most people get wrong about [topic]:
+- [Mistake 1]
+- [Mistake 2]
+- [Mistake 3]
+
+I just dropped a full breakdown on this.
+
+[Link to main video]
+
+#YourHashtags
+```
+
+**Post 2: The Insight (Friday Week 1)**
+```
+💡 [Key insight from the video]
+
+[2-3 sentences expanding on the insight]
+
+This changed how I approach [topic].
+
+Full breakdown here: [Link]
+```
+
+**Post 3: The Result/CTA (Wednesday Week 2)**
+```
+[Result or outcome statement]
+
+If you're struggling with [pain point], this will help:
+→ [Key takeaway 1]
+→ [Key takeaway 2]
+→ [Key takeaway 3]
+
+Link to watch: [Link]
+```
+
+---
+
+## Step 1: Prepare Assets
+
+**Ask the user:**
+1. "Do you have all 6 Shorts exported in vertical (9:16) format?"
+2. "Is the main video fully uploaded and scheduled on YouTube?"
+3. "Do you have the LinkedIn text posts drafted?"
+
+**Checklist:**
+- [ ] Main video uploaded (YouTube)
+- [ ] 3x Topic Shorts exported
+- [ ] 3x Extract Shorts exported
+- [ ] 3x LinkedIn text posts written
+
+---
+
+## Step 2: Schedule Cross-Platform Posts
+
+**Recommended Tools:**
+- **Later** or **Buffer** for TikTok/Instagram scheduling
+- **LinkedIn native** for LinkedIn posts
+- **YouTube Studio** for Shorts scheduling
+
+**Or Manual:** Post at these times:
+- YouTube/TikTok/IG: Post at your peak audience time (check analytics)
+- LinkedIn: Best times are Tue-Thu, 8-10am or 12pm
+
+---
+
+## Step 3: Optional Bonus Content
+
+If time allows, capture these during filming:
+
+| Content | What It Is | Platform |
+|---------|------------|----------|
+| **Behind-the-scenes photo** | You filming, setup shot | LinkedIn, Instagram Stories |
+| **B-Roll clips** | 15-second "day in the life" | Instagram Stories, TikTok |
+| **Blooper reel** | Funny outtakes | TikTok, Instagram Reels |
+
+---
+
+## Step 4: Lock Phase 5 Output
+
+```
+📌 PHASE 5 OUTPUT: DISTRIBUTION COMPLETE
+
+Main Video: [YouTube URL]
+Shorts Posted:
+- [ ] Topic Short 1: YT / TikTok / IG / LinkedIn
+- [ ] Topic Short 2: YT / TikTok / IG / LinkedIn
+- [ ] Topic Short 3: YT / TikTok / IG / LinkedIn
+- [ ] Extract Short 1: YT / TikTok / IG / LinkedIn
+- [ ] Extract Short 2: YT / TikTok / IG / LinkedIn
+- [ ] Extract Short 3: YT / TikTok / IG / LinkedIn
+
+LinkedIn Text Posts:
+- [ ] Post 1 (Hook/Pain)
+- [ ] Post 2 (Insight)
+- [ ] Post 3 (Result/CTA)
+```
+
+---
+
+## Interaction Rules
+- **If user only wants YouTube:** "You're leaving views on the table. At minimum, repost Shorts to TikTok—it's literally the same file."
+- **If user doesn't use LinkedIn:** "Skip the text posts, but still post the Shorts there. LinkedIn video is underrated."
+- **If user is overwhelmed:** "Focus on YouTube + TikTok first. Add Instagram/LinkedIn once those are habit."

package/.agent/skills/youtube-pipeline/resources/youtube-pipeline.md

@@ -0,0 +1,161 @@
+---
+description: Master orchestrator for the complete YouTube video pipeline. Chains through all 6 phases from ideation to distribution.
+---
+
+# YouTube Pipeline - Master Orchestrator
+
+> **Your Role**: You are the brutal expert YouTube strategist guiding the user through the entire video creation process. You manage state, track progress, and ensure no step is skipped.
+
+## How This Works
+This workflow chains through 6 phases:
+1. **Phase 1: Strategy Engine** - Find a mathematically proven topic (3-source validation)
+2. **Phase 2: Packaging Lab** - Engineer title & thumbnail
+3. **Phase 3: Scripting Forge** - Write a retention-optimized script
+4. **Phase 3.5: Shorts Bridge** - Create 6 Shorts (3 Topic + 3 Extract)
+5. **Phase 4: Production** - Film, edit, polish
+6. **Phase 5: Repurposing** - Distribute to all platforms
+
+You can start from Phase 1, or resume from any phase if you've already completed earlier ones.
+
+---
+
+## Entry Point
+
+Ask the user:
+"Where are we starting?
+1. **Fresh start** - Phase 1 (I have nothing yet)
+2. **I have a topic** - Phase 2 (Skip to packaging)
+3. **I have title & thumbnail** - Phase 3 (Skip to scripting)
+4. **I have a script** - Phase 4 (Skip to production)
+5. **Resume** - Pick up where we left off"
+
+---
+
+## State Tracking
+Maintain and display progress after each phase:
+
+```
+📊 PIPELINE STATUS
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+Phase 1: Strategy      [✅ COMPLETE / 🔄 IN PROGRESS / ⬜ NOT STARTED]
+Phase 2: Packaging     [✅ COMPLETE / 🔄 IN PROGRESS / ⬜ NOT STARTED]
+Phase 3: Scripting     [✅ COMPLETE / 🔄 IN PROGRESS / ⬜ NOT STARTED]
+Phase 3.5: Shorts      [✅ COMPLETE / 🔄 IN PROGRESS / ⬜ NOT STARTED]
+Phase 4: Production    [✅ COMPLETE / 🔄 IN PROGRESS / ⬜ NOT STARTED]
+Phase 5: Repurposing   [✅ COMPLETE / 🔄 IN PROGRESS / ⬜ NOT STARTED]
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+```
+
+---
+
+## Phase Handoffs
+
+### After Phase 1 → Phase 2
+Confirm outputs before proceeding:
+```
+📌 PHASE 1 OUTPUT
+- Topic: [locked]
+- Format: [locked]
+- Outlier Link: [locked]
+- Market Gap: [locked]
+```
+Then invoke: `/youtube-phase2-packaging`
+
+### After Phase 2 → Phase 3
+Confirm outputs before proceeding:
+```
+📌 PHASE 2 OUTPUT
+- Title: [locked]
+- Thumbnail Concept: [locked]
+- Thumbnail Image: [generated/prompt provided]
+- Avatar: [locked]
+- Stakes: [locked]
+```
+Then invoke: `/youtube-phase3-scripting`
+
+### After Phase 3 → Phase 3.5
+Confirm outputs before proceeding:
+```
+📌 PHASE 3 OUTPUT
+- Full Script: [locked]
+- Hook: [locked]
+- Main Points: [locked]
+- Outro CTA: [locked]
+```
+
+Ask: "Do you want to extract Shorts from this video? (Phase 3.5)"
+- If yes → invoke: `/youtube-phase3.5-shorts`
+- If no → skip to Phase 4
+
+### After Phase 3.5 → Phase 4
+Confirm outputs before proceeding:
+```
+📌 PHASE 3.5 OUTPUT (6 SHORTS)
+- Topic Short 1 (Hot Take): Script
+- Topic Short 2 (Quick Win): Script
+- Topic Short 3 (Story Hook): Script
+- Extract Short 1 (Result): Timestamp + Hook + CTA
+- Extract Short 2 (Pain): Timestamp + Hook + CTA
+- Extract Short 3 (Value): Timestamp + Hook + CTA
+```
+Then invoke: `/youtube-phase4-production`
+
+### After Phase 4 → Phase 5
+Confirm outputs before proceeding:
+```
+📌 PHASE 4 OUTPUT
+- Main Video: [exported/uploaded]
+- 6 Shorts: [exported]
+- Thumbnail: [final]
+```
+Then invoke: `/youtube-phase5-repurposing`
+
+### After Phase 5 → Complete
+```
+📌 PIPELINE COMPLETE
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+Main Video: [YouTube URL]
+Shorts: 6 posted (YouTube + TikTok + IG + LinkedIn)
+LinkedIn Posts: 3 published
+Status: FULLY DISTRIBUTED
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+```
+
+---
+
+## Phase Workflow References
+Each phase has its own detailed workflow:
+- Phase 1: `c:\CreativeOS\Creator_Command_Hub_Obsidian\📁 YouTube Brain\.agent\workflows\youtube-phase1-strategy.md`
+- Phase 2: `c:\CreativeOS\Creator_Command_Hub_Obsidian\📁 YouTube Brain\.agent\workflows\youtube-phase2-packaging.md`
+- Phase 3: `c:\CreativeOS\Creator_Command_Hub_Obsidian\📁 YouTube Brain\.agent\workflows\youtube-phase3-scripting.md`
+- Phase 3.5: `c:\CreativeOS\Creator_Command_Hub_Obsidian\📁 YouTube Brain\.agent\workflows\youtube-phase3.5-shorts.md`
+- Phase 4: `c:\CreativeOS\Creator_Command_Hub_Obsidian\📁 YouTube Brain\.agent\workflows\youtube-phase4-production.md`
+- Phase 5: `c:\CreativeOS\Creator_Command_Hub_Obsidian\📁 YouTube Brain\.agent\workflows\youtube-phase5-repurposing.md`
+
+Each workflow also reads from source knowledge:
+- Source 1: `c:\CreativeOS\Creator_Command_Hub_Obsidian\📁 YouTube Brain\📂 Processed_Notes\Workflow\Phase 1.md`
+- Source 2: `c:\CreativeOS\Creator_Command_Hub_Obsidian\📁 YouTube Brain\📂 Processed_Notes\Workflow\Phase 2.md`
+- Source 3: `c:\CreativeOS\Creator_Command_Hub_Obsidian\📁 YouTube Brain\📂 Processed_Notes\Workflow\Phase 3.md`
+- Source 4: `c:\CreativeOS\Creator_Command_Hub_Obsidian\📁 YouTube Brain\📂 Processed_Notes\Workflow\Phase 4.md`
+- Research: `c:\CreativeOS\Creator_Command_Hub_Obsidian\📁 YouTube Brain\📂 Processed_Notes\Workflow\Research.md`
+- Repurposing: `c:\CreativeOS\Creator_Command_Hub_Obsidian\📁 YouTube Brain\📂 Processed_Notes\Workflow\Repurposing.md`
+
+---
+
+## Interaction Mode Selection
+At start, ask:
+"How do you want to work?
+- **Guided** (default): I generate everything, you pick and refine
+- **Manual**: You do the work, I critique ruthlessly"
+
+User can switch modes at any step.
+
+---
+
+## Rules for the Brutal Expert
+- Never yes-man. Weak ideas get called out.
+- Demand proof for "viral" claims.
+- Teach the "why" behind every step.
+- Generate outputs for user to pick from.
+- Never let user skip steps without justification.
+- Track state across phases. Don't lose context.

package/.agent/skills/youtube-pipeline/scripts/parse_yt_studio.ps1

@@ -0,0 +1,150 @@
+# Parse YouTube Studio Inspiration HTML
+# Extracts topics and their volume indicators from exported HTML
+#
+# USAGE:
+#   1. Go to YouTube Studio → Inspiration → Search your topic
+#   2. Right-click on the results section → Inspect
+#   3. Find the parent <yta-section> element → Right-click → Copy → Copy outerHTML
+#   4. Paste into a .html file
+#   5. Run: .\parse_yt_studio.ps1 -InputFile "your_file.html"
+
+param(
+    [Parameter(Mandatory=$false)]
+    [string]$InputFile = "inspiration.html",
+    
+    [Parameter(Mandatory=$false)]
+    [string]$OutputFile = "parsed_topics.md"
+)
+
+$scriptDir = Split-Path -Parent $MyInvocation.MyCommand.Path
+$htmlPath = if ([System.IO.Path]::IsPathRooted($InputFile)) { $InputFile } else { Join-Path (Get-Location) $InputFile }
+$outputPath = if ([System.IO.Path]::IsPathRooted($OutputFile)) { $OutputFile } else { Join-Path (Get-Location) $OutputFile }
+
+if (-not (Test-Path $htmlPath)) {
+    Write-Host "Error: File not found: $htmlPath" -ForegroundColor Red
+    exit 1
+}
+
+Write-Host "Reading HTML file: $htmlPath" -ForegroundColor Cyan
+$content = Get-Content $htmlPath -Raw -Encoding UTF8
+
+if ([string]::IsNullOrWhiteSpace($content)) {
+    Write-Host "Error: File is empty or could not be read" -ForegroundColor Red
+    exit 1
+}
+
+# Pattern 1: Extract topic text from <p> tags inside topic cards
+$topicPattern = '<p\s+class="style-scope yta-content-inspiration-freeform-topic-card">\s*([^<]+?)\s*</p>'
+$topics = [regex]::Matches($content, $topicPattern) | ForEach-Object { $_.Groups[1].Value.Trim() }
+
+# Pattern 2: Extract volume indicators (Low volume, Medium volume, High volume)
+$volumePattern = '<div\s+class="additional-label[^"]*">\s*(Low volume|Medium volume|High volume|Content gap)\s*</div>'
+$volumes = [regex]::Matches($content, $volumePattern) | ForEach-Object { $_.Groups[1].Value.Trim() }
+
+Write-Host "Found $($topics.Count) topics and $($volumes.Count) volume indicators" -ForegroundColor Green
+
+# Build results - pair topics with volumes
+$results = @()
+$volumeIndex = 0
+
+for ($i = 0; $i -lt $topics.Count; $i++) {
+    $topic = $topics[$i]
+    $volume = if ($volumeIndex -lt $volumes.Count) { $volumes[$volumeIndex++] } else { "Unknown" }
+    
+    $results += [PSCustomObject]@{
+        Topic = $topic
+        Volume = $volume
+    }
+}
+
+# Sort by volume priority (High > Medium > Low)
+$volumePriority = @{
+    "High volume" = 1
+    "Medium volume" = 2
+    "Content gap" = 3
+    "Low volume" = 4
+    "Unknown" = 5
+}
+
+$sorted = $results | Sort-Object { $volumePriority[$_.Volume] }
+
+# Count by category
+$highCount = ($results | Where-Object { $_.Volume -eq "High volume" }).Count
+$medCount = ($results | Where-Object { $_.Volume -eq "Medium volume" }).Count
+$gapCount = ($results | Where-Object { $_.Volume -eq "Content gap" }).Count
+$lowCount = ($results | Where-Object { $_.Volume -eq "Low volume" }).Count
+
+# Generate markdown output
+$searchTerm = [System.IO.Path]::GetFileNameWithoutExtension($InputFile)
+$md = @"
+# YouTube Studio Inspiration - $searchTerm
+> Parsed from YouTube Studio on $(Get-Date -Format "yyyy-MM-dd HH:mm")
+
+## Summary
+- **Total Topics**: $($results.Count)
+- **High Volume**: $highCount
+- **Medium Volume**: $medCount
+- **Content Gap**: $gapCount
+- **Low Volume**: $lowCount
+
+---
+
+## 🔥 High & Medium Volume Topics (Priority)
+
+| Topic | Volume |
+|-------|--------|
+"@
+
+$highMed = $sorted | Where-Object { $_.Volume -in @("High volume", "Medium volume") }
+foreach ($item in $highMed) {
+    $md += "`n| $($item.Topic) | **$($item.Volume)** |"
+}
+
+if ($gapCount -gt 0) {
+    $md += @"
+
+---
+
+## 📊 Content Gaps (Opportunity)
+
+| Topic | Volume |
+|-------|--------|
+"@
+    $gaps = $sorted | Where-Object { $_.Volume -eq "Content gap" }
+    foreach ($item in $gaps) {
+        $md += "`n| $($item.Topic) | $($item.Volume) |"
+    }
+}
+
+$md += @"
+
+---
+
+## 📉 Low Volume Topics (For Reference)
+
+| Topic | Volume |
+|-------|--------|
+"@
+
+$low = $sorted | Where-Object { $_.Volume -eq "Low volume" }
+foreach ($item in $low) {
+    $md += "`n| $($item.Topic) | $($item.Volume) |"
+}
+
+# Write output
+$md | Out-File $outputPath -Encoding UTF8
+Write-Host "`nResults saved to: $outputPath" -ForegroundColor Green
+
+# Display top results
+Write-Host "`nTop High/Medium Volume Topics:" -ForegroundColor Yellow
+$highMed | Select-Object -First 10 | Format-Table -AutoSize
+
+# Return summary for scripting
+return @{
+    TotalTopics = $results.Count
+    HighVolume = $highCount
+    MediumVolume = $medCount
+    ContentGap = $gapCount
+    LowVolume = $lowCount
+    OutputFile = $outputPath
+}

package/.agent/workflows/LEGACY/ANTIGRAVITY_TOOLBOX.md

@@ -0,0 +1,200 @@
+# 🛠️ Antigravity Toolbox
+
+> A comprehensive reference of all tools available to the Antigravity AI agent.
+
+---
+
+## 📁 File System Operations
+
+### `view_file`
+**Purpose:** View the contents of a file from the local filesystem.
+- Supports text files (with line range selection) and some binary files (images, videos).
+- First read of a new file enforces 800 lines to understand context.
+- Max 800 lines viewable at a time.
+
+### `view_file_outline`
+**Purpose:** View the structural outline of a file.
+- Shows functions, classes, and their signatures with line ranges.
+- **Preferred first-step** for exploring file contents.
+- Only works on files, not directories.
+
+### `view_code_item`
+**Purpose:** View specific code nodes (classes/functions) by their fully qualified names.
+- Example: `Foo.bar` to view function `bar` inside class `Foo`.
+- Useful for targeted inspection of specific symbols.
+
+### `list_dir`
+**Purpose:** List contents of a directory.
+- Shows files and subdirectories with sizes and child counts.
+- Requires absolute paths.
+
+### `find_by_name`
+**Purpose:** Search for files and subdirectories using `fd`.
+- Supports glob patterns, extensions, depth limits, and exclusions.
+- Results capped at 50 matches.
+- Uses smart case and ignores gitignored files by default.
+
+### `grep_search`
+**Purpose:** Find exact pattern matches within files using `ripgrep`.
+- Returns filename, line number, and matching line content.
+- Supports regex, case-insensitive search, and glob filters.
+- Results capped at 50 matches.
+
+---
+
+## ✏️ File Editing Operations
+
+### `write_to_file`
+**Purpose:** Create new files.
+- Automatically creates parent directories if needed.
+- Supports creating artifacts for planning/documentation.
+- Can overwrite existing files if explicitly specified.
+
+### `replace_file_content`
+**Purpose:** Edit an existing file with a **single contiguous block** of changes.
+- Requires exact match of target content.
+- Use for single-area modifications.
+
+### `multi_replace_file_content`
+**Purpose:** Edit an existing file with **multiple non-contiguous** changes.
+- Specify multiple `ReplacementChunks` for scattered edits.
+- More efficient than multiple single-edit calls.
+
+---
+
+## 💻 Terminal & Command Operations
+
+### `run_command`
+**Purpose:** Propose and execute shell commands.
+- OS: Windows, Shell: PowerShell.
+- Commands require user approval unless marked safe.
+- Can run synchronously or in background.
+
+### `command_status`
+**Purpose:** Get status of a background command.
+- Check if running or done.
+- Retrieve output with configurable character count.
+
+### `send_command_input`
+**Purpose:** Send stdin to a running command or terminate it.
+- Useful for interactive REPLs and long-running processes.
+
+### `read_terminal`
+**Purpose:** Read the contents of a terminal by process ID.
+- View output from active terminal sessions.
+
+---
+
+## 🌐 Web & Browser Operations
+
+### `browser_subagent`
+**Purpose:** Spawn a browser agent to perform web interactions.
+- Click, type, navigate, resize browser.
+- Actions are **automatically recorded as WebP videos**.
+- Returns control after task completion.
+
+### `read_url_content`
+**Purpose:** Fetch content from a URL via HTTP.
+- Converts HTML to markdown.
+- No JavaScript execution or authentication.
+- For static/public content only.
+
+### `view_content_chunk`
+**Purpose:** View a specific chunk of previously fetched document content.
+- Requires prior `read_url_content` call for that document.
+
+### `search_web`
+**Purpose:** Perform a web search for a query.
+- Returns summarized information with URL citations.
+- Supports optional domain prioritization.
+
+---
+
+## 🖼️ Image Generation
+
+### `generate_image`
+**Purpose:** Generate or edit images based on text prompts.
+- Create UI mockups, assets, or visual content.
+- Can combine/edit up to 3 existing images.
+- Results saved as artifacts.
+
+---
+
+## 🔌 MCP (Model Context Protocol) Operations
+
+### `list_resources`
+**Purpose:** List available resources from an MCP server.
+- Discover what data sources are available.
+
+### `read_resource`
+**Purpose:** Retrieve contents of a specific MCP resource.
+- Access external data sources via URI.
+
+---
+
+## 📋 Task & Communication Operations
+
+### `task_boundary`
+**Purpose:** Indicate the start or update of a task.
+- Sets mode: `PLANNING`, `EXECUTION`, or `VERIFICATION`.
+- Tracks `TaskName`, `TaskStatus`, and `TaskSummary`.
+- Synchronizes with `task.md` checklist.
+
+### `notify_user`
+**Purpose:** Communicate with the user during task mode.
+- Request artifact review, ask questions, or report status.
+- Set `BlockedOnUser` if approval is needed to proceed.
+- **Only way to message user during active tasks.**
+
+---
+
+## 📊 Quick Reference Table
+
+| Tool | Category | Primary Use Case |
+|------|----------|------------------|
+| `view_file` | File System | Read file contents |
+| `view_file_outline` | File System | Explore file structure |
+| `view_code_item` | File System | Inspect specific functions/classes |
+| `list_dir` | File System | Browse directories |
+| `find_by_name` | File System | Search for files by name/pattern |
+| `grep_search` | File System | Search file contents |
+| `write_to_file` | File Editing | Create new files |
+| `replace_file_content` | File Editing | Single contiguous edit |
+| `multi_replace_file_content` | File Editing | Multiple scattered edits |
+| `run_command` | Terminal | Execute shell commands |
+| `command_status` | Terminal | Check background command status |
+| `send_command_input` | Terminal | Interact with running commands |
+| `read_terminal` | Terminal | Read terminal output |
+| `browser_subagent` | Browser | Automate browser interactions |
+| `read_url_content` | Web | Fetch webpage content |
+| `view_content_chunk` | Web | View fetched content chunks |
+| `search_web` | Web | Search the internet |
+| `generate_image` | Image | Create/edit images via AI |
+| `list_resources` | MCP | List MCP server resources |
+| `read_resource` | MCP | Read MCP resource content |
+| `task_boundary` | Task Mgmt | Track task progress |
+| `notify_user` | Communication | Message user during tasks |
+
+---
+
+## 💡 Prompting Tips
+
+1. **For file exploration:** Start with `view_file_outline` before diving into `view_file`.
+2. **For search:** Use `find_by_name` for filenames, `grep_search` for content.
+3. **For edits:** Use `replace_file_content` for single changes, `multi_replace_file_content` for multiple.
+4. **For web content:** Use `read_url_content` for static pages, `browser_subagent` for interactive sites.
+5. **For long tasks:** Request artifacts like `implementation_plan.md` for planning and `walkthrough.md` for summaries.
+6. **For commands:** Long-running commands go to background; use `command_status` to monitor.
+
+---
+
+*Last Updated: 2026-01-08*
+
+---
+
+MANDATORY: 
+1. Use view_file BEFORE every edit to get EXACT current content
+2. Fix ONE error at a time, then re-read the file
+3. Verify all variables/props exist before using them
+4. Check @[current_problems] before declaring done
+5. DO NOT PANIC if you make errors - read, understand, fix methodically