--- name: analyze-attachments description: Deep attachment analysis with extraction, importance classification, and content analysis for emails and threads. --- # Analyze Attachments Comprehensive attachment analysis including extraction, importance classification, content analysis, and intelligent cleanup. ## When to Use Use this skill when you need to: - Extract and view email attachments - Analyze document content (PDFs, contracts, invoices) - Identify important files and deadlines - Understand attachment context in conversation - Classify attachments by importance **Triggers**: English: - "Analyze attachments in ``" - "Extract attachments from ``" - "What files are attached ``" - "Show me attachments ``" - "Analyze the files in this email ``" Chinese: - "分析附件 ``" - "提取邮件附件 ``" - "这封邮件有什么附件 ``" - "查看附件内容 ``" - "附件分析 ``" ## Invocation Methods ### Method 1: Natural Language (Recommended) ```bash User: Analyze attachments AI: [Proceeds to analyze immediately] ``` ### Method 2: Slash Command ```bash User: /analyze-attachments AI: 请提供邮件 Message-ID User: AI: [Proceeds to analyze] ``` > **Note**: Can analyze attachments from single emails or entire threads. ## Operation Modes ### 1. Quick Mode (Default) ⚡ **Token Usage: Minimal** - List all attachments (filename, type, size) - Classify by importance - Identify context - ❌ Do NOT extract content **Use when**: Just need to know what's attached ### 2. Interactive Mode 🔄 **Token Usage: On-demand** - Show attachment list with importance - Ask which to extract - Extract only selected files - Analyze content **Use when**: Selective analysis needed ### 3. Auto Mode 🤖 **Token Usage: Higher** - Auto-extract high-importance files - Full content analysis - Extract deadlines and action items - Auto-cleanup temporary files **Use when**: - Email from important senders (HMRC, IRS, legal, finance) - Subject keywords (invoice, contract, urgent, tax) - User explicitly requests full analysis ## Execution Steps ### Step 1: Get Message-ID ```bash Please provide the email Message-ID 💡 Quick method (recommended): In Mail.app: Select email → Press shortcut (⌘⇧C) → Message-ID copied See README "Setup Mail Quick Action" section if not configured ``` ### Step 2: Read Email Data ```python email_data = mcp__mail__read_email(message_id) ``` ### Step 3: Analyze Attachments Check if email is part of thread: ```python thread_paths = mcp__mail__get_thread_paths(message_id) if len(thread_paths) > 1: # This is a thread # Ask user: analyze this email only or entire thread? ``` ### Step 4: Classify Attachments Importance levels: **🚨 High Importance** - Government notices (HMRC, IRS, tax documents) - Legal documents (contracts, agreements, NDAs) - Financial documents (invoices, receipts, bills) - Certificates (licenses, permits) - File type: `.pdf` from official sources **⚠️ Medium Importance** - Technical documentation - Reports and analysis - Spreadsheets and presentations - File types: `.pdf`, `.docx`, `.xlsx`, `.pptx` **💡 Low Importance** - Images (unless specified as evidence) - Signatures - Plain text files - File types: `.png`, `.jpg`, `.txt`, `.mov` ### Step 5: Choose Mode Based on Context **Auto Mode triggers**: - Sender contains: "hmrc", "irs", "tax", "court", "legal" - Subject contains: "invoice", "contract", "urgent", "deadline", "tax" - High-importance attachments detected **Interactive Mode**: - Mixed importance attachments - User seems uncertain **Quick Mode**: - Only low-importance attachments - User just browsing ### Step 6: Execute Based on Mode **Quick Mode**: Just list and classify **Interactive Mode**: ```bash Found 3 attachments: 🚨 High Importance: 1. invoice_2025.pdf (2.3 MB) ⚠️ Medium Importance: 2. report.docx (156 KB) 💡 Low Importance: 3. screenshot.png (45 KB) Would you like me to extract and analyze the high-importance file? ``` **Auto Mode**: ```bash Auto-extracting high-importance attachments... [Extract and analyze] ``` ### Step 7: Extract and Analyze (if applicable) ```python # Extract attachments attachments = mcp__mail__extract_attachments(message_id, filenames) # Analyze content for file_path in extracted_files: content = read_file(file_path) analysis = analyze_content(content) ``` ### Step 8: Cleanup **CRITICAL**: Always cleanup after analysis ```python mcp__mail__cleanup_attachments(message_ids=[message_id]) ``` ## Output Format ### Quick Mode Output ```markdown 📧 Email Attachments ### Email Information - From: [sender] - Subject: [subject] - Date: [date] ### 📎 Attachments Found ([N] files) 🚨 High Importance ([X]): 1. [filename] ([type]) - [size] - Reason: [why it's important] ⚠️ Medium Importance ([Y]): 2. [filename] ([type]) - [size] - Context: [usage] 💡 Low Importance ([Z]): 3. [filename] ([type]) - [size] ``` ### Interactive/Auto Mode Output ```markdown 📧 Attachment Analysis ### Email Information - From: [sender] - Subject: [subject] - Date: [date] ### 📎 Extracted Files Analysis 🚨 High Importance Files: **1. invoice_20250115.pdf** - Type: PDF document - Size: 2.3 MB - **Content Summary**: - Invoice number: INV-2025-001 - Amount: £1,500.00 - Due date: January 30, 2025 - Status: Pending payment - **⚠️ Action Required**: Payment due in 15 days ⚠️ Medium Importance Files: **2. project_report.docx** - Type: Word document - Size: 156 KB - **Content Summary**: - Q4 performance report - Key metrics and KPIs - **💡 Recommendation**: Review for quarterly planning 💡 Low Importance Files: **3. screenshot.png** - Type: Image - Size: 45 KB - Context: Visual reference ### 📅 Important Dates - Jan 30, 2025: Invoice payment due ### ⚠️ Action Items - [ ] Pay invoice INV-2025-001 by Jan 30 ✅ Temporary files cleaned up ``` ## Example Scenarios ### Scenario 1: Tax Notice (Auto Mode) User: Analyze attachments AI detects: - Sender: notifications@hmrc.gov.uk - Subject: "Tax Return Due 15 January" AI: Auto Mode triggered 📧 Attachment Analysis ### 📎 Extracted Files Analysis 🚨 High Importance Files: **1. tax_return_2025.pdf** - Type: PDF document - Size: 845 KB - **Content Summary**: - Tax year: 2024-2025 - Total tax due: £3,450.00 - Deadline: 15 January 2026 - Payment reference: XHJ1234567890 - **⚠️ Action Required**: Submit return and pay by Jan 15 ### 📅 Important Dates - ⚠️ Jan 15, 2026: Tax return deadline (14 days from now) ### ⚠️ Action Items - [ ] Complete tax return - [ ] Pay £3,450.00 by Jan 15 - [ ] Use reference: XHJ1234567890 ✅ Temporary files cleaned up ### Scenario 2: Mixed Attachments (Interactive Mode) User: /analyze-attachments AI: 请提供邮件 Message-ID User: AI: 📧 Email Attachments ### 📎 Attachments Found (4 files) 🚨 High Importance (1): 1. contract_draft.pdf (1.8 MB) - Legal document requiring review ⚠️ Medium Importance (2): 2. budget_2025.xlsx (245 KB) - Financial planning spreadsheet 3. project_timeline.docx (89 KB) - Schedule document 💡 Low Importance (1): 4. team_photo.jpg (1.2 MB) - Team picture 💡 Would you like me to extract and analyze the contract draft? User: Yes, please AI: [Extracts and analyzes contract] ### Scenario 3: Thread with Attachments User: Analyze attachments AI: This email is part of a thread (15 emails). Would you like to: 1. Analyze attachments from this email only 2. Analyze all attachments from the entire thread User: Option 2 AI: [Analyzes attachments across thread, shows evolution] 📎 Thread Attachment Evolution **Version History**: 1. contract_v1.pdf (Email 1) - Initial draft 2. contract_v2.pdf (Email 5) - Revised terms 3. contract_final.pdf (Email 12) - ⚠️ Final version 🚨 Most Recent: contract_final.pdf - All parties have signed - Effective date: February 1, 2025 ## Important Reminders 1. **Always cleanup** temporary files after analysis 2. **Respect user token preferences** - default to quick mode 3. **Provide context** - explain why attachments are important 4. **Track thread history** - show attachment evolution 5. **Be proactive** - suggest auto mode for important emails ## Related Skills - `/analyze-email` - For single email analysis - `/analyze-thread` - For complete thread analysis --- Source attribution: fatbobman Original project: https://github.com/fatbobman/mail-mcp-bridge/blob/638b162b26532e32fa8d8047f638537dbdfe197a/plugins/mail-attachment-analyzer/skills/analyze-attachments/SKILL.md License: MIT License information: https://opensource.org/license/mit Original upstream instructions; downloaded from Dave AI. ## Upstream license notice MIT License Copyright (c) 2025 Fatbobman (Xu Yang) Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. # Dave AI resource metadata Creator / publisher: fatbobman Original project: https://github.com/fatbobman/mail-mcp-bridge/blob/638b162b26532e32fa8d8047f638537dbdfe197a/plugins/mail-attachment-analyzer/skills/analyze-attachments/SKILL.md Attribution basis: Recovered original package and original project license at a pinned revision Fields of work: Data & Analytics, Marketing, Sales, Customer Support Roles: Customer Success Manager, Data Analyst, Email Marketer, Marketing Analyst, Support Agent Specialties: API Tools: Email (SMTP/IMAP), Gmail, Outlook Country applicability: Country not specified Classification basis: Source catalog labels Instruction SHA-256: e99a56ea0168a39df1cb5f78af1c38d213821297445254b8c67d07b3a44103b1 Metadata reviewed: 2026-10-27 This metadata supersedes earlier Dave AI catalog attribution. Original instruction and supporting files are preserved.