Cerbug45 Email Formatter

Prerequisites Check

Before using this skill, verify the environment has required capabilities: # Check if bash is available which bash # Check Python version (3.8+ required) python3 --version # Check Node.js (optional, for advanced features) node --version

Auto-Installation Script

For AI Agents: Run this installation check at first use #!/bin/bash # Email Formatter Skill - Auto Setup # This script installs required dependencies echo "🔧 Email Formatter Skill - Installation Starting..." # Create skill workspace SKILL_DIR="$HOME/.email-formatter-skill" mkdir -p "$SKILL_DIR" cd "$SKILL_DIR" # Install Python dependencies echo "📦 Installing Python packages..." pip3 install --break-system-packages --quiet \ language-tool-python \ textstat \ spacy 2>/dev/null || pip3 install --user --quiet \ language-tool-python \ textstat \ spacy # Download language model for grammar checking echo "📚 Downloading language model..." python3 -m spacy download en_core_web_sm --quiet 2>/dev/null || true # Create helper scripts directory mkdir -p "$SKILL_DIR/scripts" # Test installation if python3 -c "import language_tool_python; import textstat; import spacy" 2>/dev/null; then echo "✅ Email Formatter Skill installed successfully!" echo "📍 Installed at: $SKILL_DIR" exit 0 else echo "⚠️ Some dependencies failed. Skill will work with reduced features." exit 0 fi

Manual Installation (if auto-install fails)

# Minimal setup - no external dependencies # The skill will use built-in Python capabilities only mkdir -p ~/.email-formatter-skill/scripts echo "✅ Basic setup complete"

Verify Installation

# Quick verification python3 -c "print('Email Formatter: Ready ✅')"

Required (Always Available)

Python 3.7+ (standard library only for minimal mode) Bash shell

Optional (Enhanced Features)

language-tool-python: Advanced grammar checking textstat: Readability analysis spacy: NLP for tone detection Note for AI Agents: The skill works in degraded mode without optional dependencies. Install them only if environment allows.

Overview

This skill transforms rough email drafts into polished, professional communication by improving grammar, adjusting tone, enhancing clarity, and applying proper formatting. It handles everything from quick fixes to complete rewrites while preserving the sender's intent.

⚠️ CRITICAL SECURITY & SAFETY REQUIREMENTS

This skill MUST enforce these non-negotiable safety rules at all times:

SECURITY LEVEL: MAXIMUM - Multi-Layer Validation Required

MANDATORY PRE-PROCESSING SECURITY CHECKS: Every email MUST pass ALL security layers before any formatting occurs: Layer 1: Content Classification (BLOCK IMMEDIATELY) ❌ Illegal Activities: Fraud, scams, phishing, money laundering, tax evasion, bribery ❌ Violence & Threats: Physical threats, intimidation, stalking, doxxing, revenge threats ❌ Impersonation: Government officials, company executives, IT/support staff, law enforcement ❌ Financial Fraud: Wire transfers, cryptocurrency scams, investment fraud, Ponzi schemes ❌ Identity Theft: SSN requests, password sharing, credential phishing, fake verification ❌ Misinformation: Health fraud, election interference, conspiracy theories, fake news ❌ Child Safety: ANY content involving minors in inappropriate context ❌ Hate Speech: Racism, sexism, homophobia, religious hatred, ethnic slurs ❌ Sexual Content: Harassment, explicit content, unwanted advances, grooming ❌ Workplace Violations: Discrimination, harassment, retaliation, hostile environment ❌ Academic Fraud: Plagiarism, cheating, fake credentials, assignment ghostwriting ❌ Medical Fraud: Fake prescriptions, unlicensed advice, miracle cures, dangerous treatments ❌ Legal Violations: Contract fraud, perjury, witness tampering, obstruction ❌ Privacy Violations: Sharing private info without consent, surveillance, stalking ❌ Malware/Hacking: Phishing links, malicious attachments, system exploits ❌ Extortion: Blackmail, ransomware, threats for money, coercion Layer 2: Pattern Recognition (RED FLAGS) Scan for suspicious patterns that indicate malicious intent: Financial Red Flags: Urgent payment requests Wire transfer instructions Gift card purchases Cryptocurrency transactions "Keep this confidential" + money Bypassing normal approval process Unusual account changes Tax refund scams Inheritance scams Lottery/prize scams Authority Impersonation Red Flags: "I'm from IT/HR/Legal/Management" "CEO needs you to..." "Urgent request from [authority]" "Don't tell anyone" Bypassing email/domain verification Unusual requests from superiors Fake emergency scenarios Credential Harvesting Red Flags: "Verify your password" "Confirm your account" "Click to prevent suspension" "Unusual login detected" Links to login pages Fake security alerts Account expiration warnings Social Engineering Red Flags: Artificial urgency Emotional manipulation Too good to be true Requests for secrecy Unusual sender behavior Pressure tactics Fear-based messaging Layer 3: Sentiment & Tone Analysis (WARN OR BLOCK) ⚠️ Aggressive/Hostile: Insulting, demeaning, threatening language ⚠️ Manipulative: Guilt-tripping, gaslighting, emotional blackmail ⚠️ Coercive: Power imbalance exploitation, quid pro quo ⚠️ Deceptive: Half-truths, misleading statements, omissions ⚠️ Discriminatory: Based on protected characteristics ⚠️ Retaliatory: Punishment for protected actions Layer 4: Context Validation (VERIFY LEGITIMACY) ✓ Sender-Recipient Relationship: Does this match their normal communication? ✓ Request Reasonability: Is this a normal business request? ✓ Communication Channel: Should this be email or in-person/phone? ✓ Timing: Why is this urgent? Is urgency justified? ✓ Information Sensitivity: Should this data be in email? ✓ Authorization: Does sender have authority for this request? Layer 5: Privacy & Data Protection (GDPR/CCPA COMPLIANCE) 🔒 PII Detection: Name, address, phone, email, SSN, DOB, photos 🔒 Financial Data: Credit cards, bank accounts, tax IDs, salary info 🔒 Health Data: Medical records, diagnoses, prescriptions, HIPAA data 🔒 Credentials: Passwords, API keys, tokens, security questions 🔒 Proprietary Data: Trade secrets, confidential business info, NDA material 🔒 Children's Data: ANY data about individuals under 18 ACTION REQUIRED: If PII detected, warn user about: Email is not encrypted by default Data breach risks Regulatory compliance (GDPR, CCPA, HIPAA) Suggest secure alternatives (encrypted email, secure portal, in-person)

ZERO TOLERANCE BLOCKING - Immediate Rejection

If ANY of these detected, REFUSE IMMEDIATELY WITHOUT FORMATTING: ZERO_TOLERANCE_PATTERNS = [ # Credential Requests r'(send|give|provide).{0,20}(password|credential|login)', r'verify.{0,20}(password|account|identity)', # Financial Fraud r'wire transfer.{0,30}(urgent|immediately|today)', r'gift card.{0,20}(purchase|buy|get)', r'(bitcoin|crypto).{0,20}(send|transfer|wallet)', # Impersonation r"i'?m.{0,10}(from|with|calling from).{0,20}(IT|HR|legal|IRS|FBI)", r'(this is|i am).{0,20}(CEO|CFO|president|director)', # Threats r'(or else|otherwise).{0,30}(fire|sue|report|punish)', r'you (will|better).{0,20}(regret|pay|suffer)', # Illegal Activities r'(launder|hide|conceal).{0,20}money', r'(fake|forged|fraudulent).{0,20}(document|invoice|receipt)', # Child Safety r'(minor|child|kid|underage).{0,50}(sexual|romantic|date|meet)', # Malware/Phishing r'(click|download).{0,20}(attachment|link|file).{0,20}(urgent|immediately)', r'account.{0,20}(suspend|lock|close|expire).{0,20}(unless|until)', # Harassment r'(stupid|idiot|incompetent|worthless).{0,20}(you|employee|coworker)', r"i'?ll make sure you (never|don't|can't)", ]

Enhanced Security Response Protocol

• When prohibited content detected:
• 1. STOP - Do not process further
• 2. LOG - Record violation type (no content)
• 3. INFORM - Tell user specifically what rule was violated
• 4. EDUCATE - Explain why it's harmful/illegal
• 5. REDIRECT - Suggest legitimate alternatives
• 6. REPORT - Flag for review if severe (threats, child safety, fraud)
• Example Response Template:
• 🛑 SECURITY BLOCK: Email Formatting Refused
• REASON: [Specific violation - e.g., "Credential request detected"]
• WHY THIS IS BLOCKED:
• [Explanation - e.g., "Legitimate organizations never ask for
• passwords via email. This matches phishing attack patterns."]
• WHAT YOU SHOULD DO:
• [Alternative - e.g., "If you need to reset a password, use
• the official password reset link on the company website."]
• THIS SKILL CANNOT:
• Help with fraudulent communications
• Bypass security protocols
• Facilitate illegal activities
• Enable harassment or threats

Helper Scripts & Tools

The skill includes utility scripts for AI agents to use. Create these in ~/.email-formatter-skill/scripts/:

1. Grammar Checker (grammar_check.py)

#!/usr/bin/env python3 """ Basic grammar and spell checker Usage: python3 grammar_check.py "email text here" """ import sys import re def basic_grammar_check(text): """Basic grammar checks without external dependencies""" issues = [] # Common spelling errors typos = { 'recieve': 'receive', 'occured': 'occurred', 'seperate': 'separate', 'definately': 'definitely', 'accomodate': 'accommodate', 'tommorow': 'tomorrow', 'untill': 'until', 'truely': 'truly' } for wrong, right in typos.items(): if wrong in text.lower(): issues.append(f"Spelling: '{wrong}' → '{right}'") # Basic grammar patterns if re.search(r'\bi\s', text): # lowercase 'i' issues.append("Grammar: 'i' should be capitalized to 'I'") if re.search(r'\s{2,}', text): issues.append("Formatting: Multiple spaces detected") if re.search(r'[.!?]\s*[a-z]', text): issues.append("Grammar: Sentence should start with capital letter") # Double punctuation if re.search(r'[.!?]{2,}', text): issues.append("Punctuation: Multiple punctuation marks") return issues if __name__ == "__main__": if len(sys.argv) < 2: print("Usage: python3 grammar_check.py 'text'") sys.exit(1) text = sys.argv[1] issues = basic_grammar_check(text) if issues: for issue in issues: print(f"⚠️ {issue}") else: print("✅ No basic issues found")

2. Tone Analyzer (tone_analyzer.py)

#!/usr/bin/env python3 """ Analyze email tone Usage: python3 tone_analyzer.py "email text" """ import sys import re def analyze_tone(text): """Detect tone indicators in email text""" # Formal indicators formal_words = ['pursuant', 'hereby', 'aforementioned', 'regarding', 'sincerely', 'respectfully', 'cordially'] # Casual indicators casual_words = ['hey', 'gonna', 'wanna', 'yeah', 'yep', 'nope', 'btw', 'fyi', 'lol', 'omg', 'tbh'] # Aggressive indicators aggressive_words = ['immediately', 'must', 'unacceptable', 'ridiculous', 'obviously', 'clearly', 'need to', 'have to'] # Polite indicators polite_words = ['please', 'kindly', 'would you', 'could you', 'appreciate', 'thank', 'grateful'] text_lower = text.lower() formal_count = sum(1 for w in formal_words if w in text_lower) casual_count = sum(1 for w in casual_words if w in text_lower) aggressive_count = sum(1 for w in aggressive_words if w in text_lower) polite_count = sum(1 for w in polite_words if w in text_lower) # Exclamation marks exclamations = len(re.findall(r'!', text)) # ALL CAPS detection caps_words = len(re.findall(r'\b[A-Z]{2,}\b', text)) # Determine primary tone tones = [] if formal_count >= 2: tones.append("FORMAL") if casual_count >= 2: tones.append("CASUAL") if aggressive_count >= 2 or caps_words >= 2: tones.append("AGGRESSIVE") if polite_count >= 2: tones.append("POLITE") if exclamations >= 3: tones.append("ENTHUSIASTIC/URGENT") if not tones: tones.append("NEUTRAL") return { 'primary_tone': tones[0], 'all_tones': tones, 'formal_score': formal_count, 'casual_score': casual_count, 'aggressive_score': aggressive_count, 'polite_score': polite_count, 'exclamations': exclamations, 'caps_words': caps_words } if __name__ == "__main__": if len(sys.argv) < 2: print("Usage: python3 tone_analyzer.py 'text'") sys.exit(1) result = analyze_tone(sys.argv[1]) print(f"📊 Primary Tone: {result['primary_tone']}") print(f"🎯 All Tones: {', '.join(result['all_tones'])}") print(f"📈 Scores - Formal:{result['formal_score']} Casual:{result['casual_score']} " f"Aggressive:{result['aggressive_score']} Polite:{result['polite_score']}") if result['aggressive_score'] >= 2: print("⚠️ WARNING: Email may sound aggressive") if result['exclamations'] >= 3: print("⚠️ WARNING: Too many exclamation marks") if result['caps_words'] >= 2: print("⚠️ WARNING: Excessive capitalization detected")

3. Readability Scorer (readability.py)

#!/usr/bin/env python3 """ Calculate email readability Usage: python3 readability.py "email text" """ import sys import re def count_syllables(word): """Simple syllable counter""" word = word.lower() vowels = 'aeiouy' syllable_count = 0 previous_was_vowel = False for char in word: is_vowel = char in vowels if is_vowel and not previous_was_vowel: syllable_count += 1 previous_was_vowel = is_vowel # Adjust for silent 'e' if word.endswith('e'): syllable_count -= 1 # Every word has at least one syllable if syllable_count == 0: syllable_count = 1 return syllable_count def flesch_reading_ease(text): """Calculate Flesch Reading Ease score""" sentences = len(re.findall(r'[.!?]+', text)) or 1 words = len(text.split()) syllables = sum(count_syllables(word) for word in text.split()) if words == 0: return 0 score = 206.835 - 1.015 * (words / sentences) - 84.6 * (syllables / words) return round(score, 1) def analyze_readability(text): """Analyze email readability""" words = text.split() sentences = len(re.findall(r'[.!?]+', text)) or 1 avg_word_length = sum(len(w) for w in words) / len(words) if words else 0 avg_sentence_length = len(words) / sentences flesch_score = flesch_reading_ease(text) # Interpret score if flesch_score >= 90: level = "Very Easy (5th grade)" elif flesch_score >= 80: level = "Easy (6th grade)" elif flesch_score >= 70: level = "Fairly Easy (7th grade)" elif flesch_score >= 60: level = "Standard (8-9th grade)" elif flesch_score >= 50: level = "Fairly Difficult (10-12th grade)" elif flesch_score >= 30: level = "Difficult (College)" else: level = "Very Difficult (Graduate)" return { 'flesch_score': flesch_score, 'level': level, 'avg_word_length': round(avg_word_length, 1), 'avg_sentence_length': round(avg_sentence_length, 1), 'total_words': len(words), 'total_sentences': sentences } if __name__ == "__main__": if len(sys.argv) < 2: print("Usage: python3 readability.py 'text'") sys.exit(1) result = analyze_readability(sys.argv[1]) print(f"📖 Flesch Reading Ease: {result['flesch_score']}") print(f"📚 Reading Level: {result['level']}") print(f"📊 Stats: {result['total_words']} words, {result['total_sentences']} sentences") print(f"📏 Avg: {result['avg_word_length']} chars/word, {result['avg_sentence_length']} words/sentence") # Recommendations if result['flesch_score'] < 60: print("💡 TIP: Simplify language for better clarity") if result['avg_sentence_length'] > 20: print("💡 TIP: Break long sentences into shorter ones")

4. Security Scanner (security_scan.py)

#!/usr/bin/env python3 """ ULTRA-SECURE Email Scanner - Multi-Layer Threat Detection Usage: python3 security_scan.py "email text" Exit codes: 0=safe, 1=warning, 2=critical_block, 3=report_required """ import sys import re import json from datetime import datetime class SecurityScanner: """Military-grade email security scanner""" def __init__(self): self.threat_level = 0 # 0=safe, 1=warning, 2=critical, 3=report self.violations = [] self.warnings = [] def scan(self, text): """Run all security checks""" # Layer 1: Zero Tolerance Patterns self.check_zero_tolerance(text) # Layer 2: Financial Fraud self.check_financial_fraud(text) # Layer 3: Impersonation self.check_impersonation(text) # Layer 4: Credential Harvesting self.check_credential_harvesting(text) # Layer 5: Threats & Violence self.check_threats(text) # Layer 6: Harassment & Discrimination self.check_harassment(text) # Layer 7: Privacy & PII self.check_privacy_violations(text) # Layer 8: Social Engineering self.check_social_engineering(text) # Layer 9: Child Safety self.check_child_safety(text) # Layer 10: Malicious Patterns self.check_malicious_patterns(text) return self.generate_report() def check_zero_tolerance(self, text): """Critical patterns that immediately block""" text_lower = text.lower() critical_patterns = [ # Credentials (r'(send|give|provide|email).{0,30}(password|pwd|credential|login|passphrase)', 'CREDENTIAL_REQUEST', 3), (r'verify.{0,20}(password|account|identity|credential)', 'FAKE_VERIFICATION', 3), (r'(username|user id).{0,20}(and|&|\\+).{0,20}password', 'CREDENTIAL_PAIR_REQUEST', 3), # Financial (r'wire transfer.{0,30}(urgent|immediate|asap|now|today)', 'URGENT_WIRE_TRANSFER', 3), (r'(gift card|itunes|steam|amazon card).{0,30}(buy|purchase|get|send)', 'GIFT_CARD_SCAM', 3), (r'(bitcoin|btc|crypto|ethereum|eth).{0,30}(wallet|address|send|transfer)', 'CRYPTO_SCAM', 3), (r'(bank account|routing number|swift code).{0,30}(provide|send|give)', 'BANKING_INFO_REQUEST', 3), # Impersonation (r"i'?m.{0,10}(from|with|calling from).{0,30}(IT|support|tech|help desk)", 'IT_IMPERSONATION', 3), (r"(this is|i am|i'm).{0,20}(CEO|CFO|president|director|executive)", 'EXECUTIVE_IMPERSONATION', 3), (r"(IRS|FBI|police|government|immigration).{0,30}(contact|reach out|notice)", 'AUTHORITY_IMPERSONATION', 3), # Threats (r'(or else|otherwise).{0,30}(fire|terminate|sue|report|arrest)', 'THREAT_DETECTED', 3), (r"(you|i)'?(ll| will).{0,30}(regret|pay|suffer|sorry)", 'THREAT_LANGUAGE', 3), # Child Safety (r'(child|minor|kid|teen|underage).{0,50}(meet|date|relationship|romantic)', 'CHILD_SAFETY_VIOLATION', 3), # Malware (r'(click|open|download).{0,20}(attachment|link|file).{0,20}(urgent|expire|suspend)', 'PHISHING_LINK', 3), ] for pattern, violation_type, severity in critical_patterns: if re.search(pattern, text_lower): self.add_violation(violation_type, severity, pattern) def check_financial_fraud(self, text): """Detect financial scam patterns""" text_lower = text.lower() fraud_indicators = [ (r'(won|winner|prize|lottery).{0,30}(\$|dollar|money|claim)', 'LOTTERY_SCAM', 2), (r'(inheritance|beneficiary|estate).{0,50}(million|claim|transfer)', 'INHERITANCE_SCAM', 2), (r'(tax|irs).{0,30}(refund|owe|pay immediately)', 'TAX_SCAM', 2), (r'(invoice|payment).{0,20}(overdue|urgent|immediate|final notice)', 'FAKE_INVOICE', 2), (r'(suspended|frozen|locked).{0,30}account', 'ACCOUNT_SUSPENSION_SCAM', 2), (r'(refund|reimbursement).{0,30}(click|verify|confirm)', 'REFUND_SCAM', 2), (r'(investment|opportunity|profit).{0,50}(guaranteed|risk-free|double)', 'INVESTMENT_FRAUD', 2), ] for pattern, violation_type, severity in fraud_indicators: if re.search(pattern, text_lower): self.add_violation(violation_type, severity, pattern) def check_impersonation(self, text): """Detect impersonation attempts""" text_lower = text.lower() impersonation_patterns = [ (r'(on behalf of|representing).{0,30}(company|organization|government)', 'UNAUTHORIZED_REPRESENTATION', 2), (r"(i'?m|this is).{0,20}(calling|writing|reaching out).{0,20}(from|regarding)", 'SUSPICIOUS_INTRODUCTION', 1), (r'(verify|confirm).{0,20}(you are|your identity|who you are)', 'IDENTITY_VERIFICATION_REQUEST', 2), ] for pattern, violation_type, severity in impersonation_patterns: if re.search(pattern, text_lower): self.add_violation(violation_type, severity, pattern) def check_credential_harvesting(self, text): """Detect credential theft attempts""" text_lower = text.lower() patterns = [ (r'(account|access).{0,30}(expire|suspend|lock|disable)', 'FAKE_EXPIRATION', 2), (r'(security|unusual|suspicious).{0,30}activity', 'FAKE_SECURITY_ALERT', 2), (r'(update|verify|confirm).{0,30}(payment|billing) (method|information)', 'PAYMENT_INFO_PHISHING', 2), (r'(reset|recover|change).{0,20}password.{0,20}(click|link|here)', 'PASSWORD_RESET_SCAM', 2), ] for pattern, violation_type, severity in patterns: if re.search(pattern, text_lower): self.add_violation(violation_type, severity, pattern) def check_threats(self, text): """Detect threats and violent language""" text_lower = text.lower() threat_words = [ 'kill', 'hurt', 'harm', 'destroy', 'eliminate', 'punish', 'revenge', 'retaliate', 'get back at', 'make you pay' ] for word in threat_words: if word in text_lower: self.add_violation('THREAT_LANGUAGE', 3, f"Threat word: {word}") # Physical threat patterns if re.search(r'(come after|find you|know where you)', text_lower): self.add_violation('PHYSICAL_THREAT', 3, 'Physical threat implied') def check_harassment(self, text): """Detect harassment and hostile language""" text_lower = text.lower() hostile_words = [ 'stupid', 'idiot', 'moron', 'incompetent', 'worthless', 'pathetic', 'useless', 'loser', 'failure', 'trash' ] count = sum(1 for word in hostile_words if word in text_lower) if count >= 2: self.add_violation('HARASSMENT', 2, f'{count} hostile terms detected') elif count == 1: self.add_warning('POTENTIALLY_HOSTILE', 'Hostile language detected') # Discriminatory patterns protected_characteristics = [ (r'(because|since).{0,20}(you\'?re|you are).{0,20}(woman|female|girl)', 'GENDER_DISCRIMINATION'), (r'(because|since).{0,20}(you\'?re|you are).{0,20}(old|young|age)', 'AGE_DISCRIMINATION'), (r'(people like you|your kind|you people)', 'DISCRIMINATORY_LANGUAGE'), ] for pattern, violation_type in protected_characteristics: if re.search(pattern, text_lower): self.add_violation(violation_type, 3, pattern) def check_privacy_violations(self, text): """Detect PII and privacy issues""" # SSN pattern if re.search(r'\b\d{3}-\d{2}-\d{4}\b', text): self.add_violation('SSN_DETECTED', 2, 'Social Security Number found') # Credit card pattern if re.search(r'\b\d{4}[\s-]?\d{4}[\s-]?\d{4}[\s-]?\d{4}\b', text): self.add_violation('CREDIT_CARD_DETECTED', 2, 'Credit card number found') # Email addresses (multiple) emails = re.findall(r'\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Z|a-z]{2,}\b', text) if len(emails) > 3: self.add_warning('MULTIPLE_EMAILS', f'{len(emails)} email addresses found') # Phone numbers (multiple) phones = re.findall(r'\b\d{3}[-.]?\d{3}[-.]?\d{4}\b', text) if len(phones) > 2: self.add_warning('MULTIPLE_PHONES', f'{len(phones)} phone numbers found') # Home address pattern if re.search(r'\d+\s+\w+\s+(street|st|avenue|ave|road|rd|drive|dr)', text.lower()): self.add_warning('ADDRESS_DETECTED', 'Physical address found') def check_social_engineering(self, text): """Detect social engineering tactics""" text_lower = text.lower() # Urgency indicators urgency_words = ['urgent', 'immediate', 'asap', 'right now', 'immediately', 'emergency', 'critical', 'time-sensitive'] urgency_count = sum(1 for word in urgency_words if word in text_lower) if urgency_count >= 3: self.add_violation('ARTIFICIAL_URGENCY', 2, f'{urgency_count} urgency indicators') elif urgency_count >= 2: self.add_warning('URGENCY_DETECTED', 'Multiple urgency indicators') # Secrecy requests if re.search(r"(don't tell|keep (this )?secret|confidential|between us)", text_lower): self.add_violation('SECRECY_REQUEST', 2, 'Requesting secrecy') # Authority bypass if re.search(r'(bypass|skip|ignore).{0,20}(normal|usual|standard) (process|procedure)', text_lower): self.add_violation('PROCESS_BYPASS', 2, 'Attempting to bypass normal procedures') # Too good to be true if re.search(r'(free|win|won|winner|selected|chosen).{0,30}(prize|money|gift|\$)', text_lower): self.add_warning('TOO_GOOD_TO_BE_TRUE', 'Unrealistic offer detected') def check_child_safety(self, text): """Critical: Child safety violations""" text_lower = text.lower() child_terms = ['child', 'minor', 'kid', 'teen', 'teenager', 'underage', 'student', 'pupil'] inappropriate_context = ['date', 'dating', 'romantic', 'relationship', 'meet in person', 'alone', 'secret', 'don\'t tell', 'special friend'] has_child_term = any(term in text_lower for term in child_terms) has_inappropriate = any(term in text_lower for term in inappropriate_context) if has_child_term and has_inappropriate: self.add_violation('CHILD_SAFETY_CRITICAL', 3, 'Child safety violation - REPORT REQUIRED') def check_malicious_patterns(self, text): """Detect malware and hacking patterns""" text_lower = text.lower() malicious_patterns = [ (r'(click|open).{0,20}attachment.{0,20}(urgent|important|invoice)', 'MALICIOUS_ATTACHMENT', 2), (r'(download|install|run).{0,20}(software|program|tool|update)', 'UNAUTHORIZED_SOFTWARE', 2), (r'(disable|turn off).{0,20}(antivirus|firewall|security)', 'SECURITY_BYPASS', 3), (r'(admin|administrator|root).{0,20}(access|password|privileges)', 'PRIVILEGE_ESCALATION', 3), ] for pattern, violation_type, severity in malicious_patterns: if re.search(pattern, text_lower): self.add_violation(violation_type, severity, pattern) def add_violation(self, violation_type, severity, pattern): """Record a security violation""" self.violations.append({ 'type': violation_type, 'severity': severity, 'pattern': pattern, 'timestamp': datetime.now().isoformat() }) if severity > self.threat_level: self.threat_level = severity def add_warning(self, warning_type, message): """Record a warning""" self.warnings.append({ 'type': warning_type, 'message': message }) if self.threat_level == 0: self.threat_level = 1 def generate_report(self): """Generate security scan report""" return { 'threat_level': self.threat_level, 'status': self.get_status(), 'violations': self.violations, 'warnings': self.warnings, 'summary': self.get_summary() } def get_status(self): """Get security status""" if self.threat_level >= 3: return 'CRITICAL_BLOCK_AND_REPORT' elif self.threat_level == 2: return 'BLOCK' elif self.threat_level == 1: return 'WARNING' else: return 'SAFE' def get_summary(self): """Get human-readable summary""" if self.threat_level >= 3: return f"🚨 CRITICAL: {len(self.violations)} severe violations detected. DO NOT SEND. REPORT REQUIRED." elif self.threat_level == 2: return f"🛑 BLOCKED: {len(self.violations)} violations detected. Cannot format this email." elif self.threat_level == 1: return f"⚠️ WARNING: {len(self.warnings)} potential issues detected. Review carefully." else: return "✅ No security issues detected." if __name__ == "__main__": if len(sys.argv) < 2: print("Usage: python3 security_scan.py 'email text'") sys.exit(1) scanner = SecurityScanner() report = scanner.scan(sys.argv[1]) # Print report print(f"\n{'='*60}") print(f"SECURITY SCAN REPORT") print(f"{'='*60}") print(f"Status: {report['status']}") print(f"Threat Level: {report['threat_level']}/3") print(f"\n{report['summary']}\n") if report['violations']: print("VIOLATIONS:") for v in report['violations']: severity_icon = "🚨" if v['severity'] >= 3 else "🛑" print(f" {severity_icon} {v['type']}") print(f" Pattern: {v['pattern'][:50]}...") if report['warnings']: print("\nWARNINGS:") for w in report['warnings']: print(f" ⚠️ {w['type']}: {w['message']}") print(f"\n{'='*60}\n") # Return appropriate exit code sys.exit(report['threat_level'])

Usage Workflow for AI Agents

MANDATORY SECURITY-FIRST WORKFLOW: #!/bin/bash # Email Formatter - Secure Processing Pipeline # This workflow is REQUIRED for every email formatting request set -e # Exit on any error EMAIL_TEXT="$1" TEMP_DIR="/tmp/email-formatter-$$" mkdir -p "$TEMP_DIR" echo "🔒 Starting Secure Email Processing Pipeline..." echo "================================================" # STEP 1: PRE-FLIGHT SECURITY SCAN (CRITICAL) echo "Step 1/7: Running security scan..." python3 ~/.email-formatter-skill/scripts/security_scan.py "$EMAIL_TEXT" SECURITY_EXIT=$? if [ $SECURITY_EXIT -eq 3 ]; then echo "" echo "🚨🚨🚨 CRITICAL SECURITY VIOLATION 🚨🚨🚨" echo "This email contains SEVERE violations that must be reported." echo "Formatting REFUSED. Potential illegal activity detected." echo "" echo "ACTIONS REQUIRED:" echo "1. Do NOT send this email" echo "2. Document the incident" echo "3. Report to appropriate authorities if applicable" echo "4. Inform user of violation" exit 3 elif [ $SECURITY_EXIT -eq 2 ]; then echo "" echo "🛑 SECURITY BLOCK" echo "This email violates safety policies and cannot be formatted." echo "Review the security report above for specific violations." echo "" echo "SUGGESTED ACTIONS:" echo "1. Identify the specific violation" echo "2. Explain to user why it's blocked" echo "3. Suggest legitimate alternatives" echo "4. Offer to help rewrite with appropriate content" exit 2 elif [ $SECURITY_EXIT -eq 1 ]; then echo "" echo "⚠️ SECURITY WARNING" echo "Potential issues detected. Proceeding with caution..." echo "Will re-scan after formatting to ensure no issues introduced." echo "" fi # STEP 2: CONTENT ANALYSIS echo "" echo "Step 2/7: Analyzing content..." echo "$EMAIL_TEXT" > "$TEMP_DIR/original.txt" # Word count WORD_COUNT=$(echo "$EMAIL_TEXT" | wc -w) echo " 📊 Word count: $WORD_COUNT" if [ $WORD_COUNT -gt 500 ]; then echo " ⚠️ Email is very long. Consider breaking into multiple emails." fi # STEP 3: TONE ANALYSIS echo "" echo "Step 3/7: Analyzing tone..." python3 ~/.email-formatter-skill/scripts/tone_analyzer.py "$EMAIL_TEXT" > "$TEMP_DIR/tone.txt" cat "$TEMP_DIR/tone.txt" # Check if tone is aggressive if grep -q "AGGRESSIVE" "$TEMP_DIR/tone.txt"; then echo "" echo " ⚠️ AGGRESSIVE TONE DETECTED" echo " Recommendation: Suggest user wait 24 hours before sending" echo " Offer to rewrite in professional, constructive tone" echo "" read -p " Continue anyway? (yes/no): " CONTINUE if [ "$CONTINUE" != "yes" ]; then echo " Formatting cancelled by tone check." exit 1 fi fi # STEP 4: GRAMMAR CHECK echo "" echo "Step 4/7: Checking grammar..." python3 ~/.email-formatter-skill/scripts/grammar_check.py "$EMAIL_TEXT" # STEP 5: READABILITY ANALYSIS echo "" echo "Step 5/7: Analyzing readability..." python3 ~/.email-formatter-skill/scripts/readability.py "$EMAIL_TEXT" > "$TEMP_DIR/readability.txt" cat "$TEMP_DIR/readability.txt" # STEP 6: FORMAT THE EMAIL echo "" echo "Step 6/7: Formatting email..." echo " [Agent applies formatting rules based on guidelines]" # ... AI agent performs formatting here ... # FORMATTED_EMAIL="..." # STEP 7: POST-FORMATTING SECURITY RE-SCAN (CRITICAL) echo "" echo "Step 7/7: Final security validation..." python3 ~/.email-formatter-skill/scripts/security_scan.py "$FORMATTED_EMAIL" FINAL_SECURITY=$? if [ $FINAL_SECURITY -ne 0 ]; then echo "" echo "🚨 POST-FORMATTING SECURITY FAILURE" echo "Formatted version introduced security issues!" echo "This should never happen - formatting logic has a bug." echo "Reverting to original and blocking output." exit 2 fi # SUCCESS echo "" echo "✅ All security checks passed" echo "✅ Email formatted successfully" echo "" echo "================================================" echo "FORMATTED EMAIL READY FOR REVIEW" echo "================================================" # Cleanup rm -rf "$TEMP_DIR"

Critical Enforcement Rules for AI Agents

NEVER SKIP SECURITY SCANS: Security scan must run BEFORE any formatting Security scan must run AFTER formatting Both scans must pass (exit 0) for email to be delivered Exit codes 1, 2, 3 MUST stop the pipeline DOUBLE-CHECK MECHANISM: # Before presenting formatted email to user def final_validation(formatted_email): # Re-run all security checks security_clear = run_security_scan(formatted_email) if not security_clear: # NEVER show formatted email return { 'status': 'BLOCKED', 'message': 'Formatting introduced security issues', 'action': 'Report bug in formatting logic' } # Additional checks if contains_pii(formatted_email): return { 'status': 'WARNING', 'message': 'PII detected in formatted email', 'action': 'Warn user about sending sensitive data via email' } return { 'status': 'APPROVED', 'formatted_email': formatted_email } Logging & Audit Trail (Privacy-Safe): # Log violations only (NO content) def log_security_event(violation_type, severity, timestamp): """ Log security events for monitoring NEVER log actual email content """ log_entry = { 'timestamp': timestamp, 'violation_type': violation_type, 'severity': severity, 'action_taken': 'BLOCKED', 'content': '[REDACTED]' # Never log content } # Append to secure audit log # This helps improve security detection

When to Use This Skill

Use this skill when the user needs to: Fix grammar, spelling, and punctuation in emails Adjust tone (make more formal, casual, friendly, or assertive) Structure messy drafts into clear, organized messages Convert bullet points or notes into complete emails Add professional greetings and closings Improve clarity and conciseness Prepare important messages for executives, clients, or stakeholders

Core Principles

Preserve Intent: Never change the core message or facts - only improve how it's communicated Match Context: Adjust formality based on recipient and situation Enhance Clarity: Remove ambiguity while maintaining natural voice Professional Standard: Apply business writing best practices Cultural Awareness: Consider professional norms and communication styles

Step 1: Analyze the Draft

Before formatting, assess: Recipient relationship: Boss, colleague, client, vendor, team, external? Purpose: Request, update, introduction, follow-up, feedback, apology? Tone needed: Formal, semi-formal, casual, friendly, assertive, diplomatic? Urgency: Routine, important, urgent, sensitive? Current issues: Grammar errors, unclear structure, wrong tone, missing context?

Step 2: Apply Improvements

Grammar & Mechanics: Fix spelling, punctuation, and grammatical errors Correct subject-verb agreement and tense consistency Remove run-on sentences and fragments Fix comma splices and misplaced modifiers Structure & Organization: Standard Email Structure: 1. Greeting (appropriate to relationship) 2. Opening (context or pleasantry) 3. Purpose statement (clear and direct) 4. Body (organized by topic, use paragraphs/bullets) 5. Call to action (if needed) 6. Closing (polite sign-off) 7. Signature Tone Adjustments: Too Casual → Professional: Before: "Hey! Just wanted to check if u got my last email lol" After: "Hi Sarah, I wanted to follow up on my previous email from Tuesday. Please let me know if you need any additional information." Too Formal → Friendly: Before: "I am writing to inquire whether you have completed the aforementioned task." After: "Hi John, I wanted to check in on the status of the marketing report. How's it coming along?" Too Aggressive → Diplomatic: Before: "You need to fix this immediately. This is unacceptable." After: "I noticed an issue that requires urgent attention. Could we prioritize resolving this today? I'm happy to help if needed." Clarity Enhancements: Replace vague phrases with specific language Break long paragraphs into digestible chunks Use bullet points for lists or multiple items Add context where assumed knowledge might be missing Remove redundancy and filler words

Step 3: Polish Details

Subject Line (if provided or needed): Keep it under 50 characters Make it specific and actionable Use sentence case (not all caps) Examples: "Q1 Budget Review Meeting - March 15" "Quick question about project timeline" "Following up: Website redesign proposal" Greetings: Formal: "Dear Dr. Smith," or "Dear Hiring Manager," Professional: "Hi Jennifer," or "Hello Team," Casual: "Hey Alex," or "Hi everyone," Closings: Formal: "Sincerely," "Respectfully," "Best regards," Professional: "Best," "Thanks," "Looking forward to hearing from you," Casual: "Cheers," "Talk soon," "Have a great day," Signature Block: Best regards, [Name] [Title] [Company] [Contact Info - if external]

1. Request Email

• Structure:
• Greeting
• Context (why you're writing)
• Specific request
• Deadline or timeframe (if applicable)
• Offer of additional info
• Thanks
• Closing

2. Follow-Up Email

• Structure:
• Reference previous communication
• Polite reminder of action needed
• Make it easy to respond
• Maintain friendly tone
• Closing

3. Bad News Email

• Structure:
• Direct but empathetic opening
• Clear explanation
• Acknowledge impact
• Offer alternatives or next steps
• End on positive note if possible

4. Introduction Email

• Structure:
• Who you are and connection
• Purpose of introduction
• What you're offering/requesting
• Call to action
• Professional closing

DO:

✅ Keep emails concise (under 200 words when possible) ✅ Use active voice ("I will send" vs "It will be sent") ✅ Break up text with white space ✅ Put most important info in first paragraph ✅ Proofread for typos and auto-correct errors ✅ Use "Reply All" judiciously ✅ Include clear next steps or calls to action ✅ Match the sender's energy level

DON'T:

❌ Use all caps (seems like shouting) ❌ Overuse exclamation marks ❌ Include multiple topics in one email (if complex) ❌ Use jargon with external recipients ❌ Write when emotional (flag if email seems angry) ❌ Assume tone translates (sarcasm, humor can fail) ❌ Forget attachments referenced in text ❌ Change factual content or commitments

Tone Guide

Formal (executives, clients, first contact): Complete sentences Professional vocabulary Proper titles and full names "I would appreciate" vs "Can you" "Please let me know" vs "Let me know" Semi-Formal (colleagues, regular contacts): Conversational but professional Contractions are fine First names acceptable "Could you" vs "Can you" Friendly but respectful Casual (close colleagues, internal teams): Relaxed language Contractions and informal phrases Quick greetings Can be brief Emoji okay if culturally appropriate

Quality Checklist

Before presenting the formatted email, verify: SECURITY FIRST: Content passes all safety requirements No prohibited content: Checked against all safety rules above Legal compliance: No fraudulent, harassing, or illegal content Ethical standards: Message is honest and appropriate Grammar and spelling are correct Tone matches relationship and context Structure is clear and logical Key information is easy to find Call to action is clear (if needed) Opening and closing are appropriate No ambiguity or confusion Length is appropriate (concise but complete) Professional formatting applied Original intent is preserved Privacy check: No sensitive data exposed inappropriately Reputation check: Sender won't regret sending this

Red Flag Detection

Always scan for these warning signs: Requests for money, credentials, or personal information Urgency combined with financial requests Impersonation language ("I'm calling from...", "This is [authority]...") Threats or ultimatums Asking recipient to keep communication secret Bypassing normal procedures Inconsistent sender information Requests to click suspicious links Grammar/spelling errors in supposedly official communication Too-good-to-be-true offers Emotional manipulation tactics Discriminatory language False information Hostile or aggressive tone toward protected groups

Incident Response Protocol

• When Critical Violations Detected (Threat Level 3):
• IMMEDIATE ACTIONS:
• 1. BLOCK - Refuse to format email
• 2. DOCUMENT - Record violation type, timestamp
• 3. NOTIFY - Inform user of specific violation
• 4. EDUCATE - Explain why it's harmful/illegal
• 5. REDIRECT - Suggest legitimate alternatives
• 6. REPORT - Flag for review if:
• - Child safety violations
• - Credible threats of violence
• - Large-scale fraud attempts
• - Illegal activities
• Response Template for Critical Violations:
• 🚨 CRITICAL SECURITY VIOLATION DETECTED
• VIOLATION TYPE: [Specific type - e.g., "Credential Phishing Attempt"]
• SEVERITY: CRITICAL - This email cannot be formatted
• WHAT WAS DETECTED:
• [Specific pattern - e.g., "Email requests password and account
• credentials, matching known phishing attack patterns"]
• WHY THIS IS SERIOUS:
• [Impact - e.g., "This could lead to:
• Identity theft
• Unauthorized account access
• Financial fraud
• Legal liability for sender"]
• WHAT YOU SHOULD KNOW:
• Legitimate organizations NEVER ask for passwords via email
• This pattern is used in 95% of credential phishing attacks
• Sending this email could violate anti-fraud laws
• RECOMMENDED ACTIONS:
• 1. If you received a similar email: Report it as phishing
• 2. If you need password help: Use official password reset tools
• 3. If suspicious: Contact IT/security team directly
• ALTERNATIVE APPROACH:
• [Legitimate way to accomplish goal if applicable]
• ---
• This email has been blocked to protect you and recipients.
• For questions about this decision, review security guidelines.

Security Metrics & Monitoring

Track these metrics (content-free): SECURITY_METRICS = { 'total_scans': 0, 'threats_blocked': { 'level_1_warnings': 0, 'level_2_blocks': 0, 'level_3_critical': 0 }, 'violation_types': { 'phishing': 0, 'fraud': 0, 'threats': 0, 'harassment': 0, 'impersonation': 0, 'pii_exposure': 0, 'malware': 0, 'child_safety': 0 }, 'false_positives_reported': 0, 'scan_performance_ms': [] } Regular Security Audits: Review blocked emails (patterns only, no content) Update detection patterns based on new threats Tune sensitivity to reduce false positives Improve educational messages Add new threat categories as they emerge

Privacy & Data Protection Compliance

GDPR/CCPA/HIPAA Compliance: DATA_PROTECTION_RULES = { 'data_minimization': 'Process only what's needed for formatting', 'purpose_limitation': 'Use data ONLY for formatting, nothing else', 'storage_limitation': 'Delete immediately after processing', 'accuracy': 'Don't modify factual content', 'integrity': 'Secure processing, encrypted if possible', 'confidentiality': 'Treat all emails as confidential', 'accountability': 'Log violations (not content) for audit' } User Rights: Right to explanation: User can ask why email was blocked Right to appeal: User can dispute false positives Right to deletion: No content retention Right to transparency: Clear security policies

Continuous Improvement

Security Pattern Updates: # Periodically update threat patterns # Add new scam patterns as they emerge # Improve detection accuracy # Reduce false positives # Example: Update security_scan.py git pull origin main # Get latest threat patterns python3 -m pytest tests/test_security.py # Verify Feedback Loop: Collect anonymized violation statistics Identify common false positives Refine pattern matching Add context-aware detection Improve user education messages

Output Format

• Present the formatted email like this:
• **Subject:** [Subject line if relevant]
• [Formatted email body]
• ---
• **Changes Made:**
• [Brief list of key improvements]
• [E.g., "Corrected grammar errors"]
• [E.g., "Adjusted tone from casual to professional"]
• [E.g., "Restructured for clarity"]

Special Considerations

Security Edge Cases: Example 1 - Phishing Attempt: User Input: "Format this: Hi, I'm from IT. Send me your password to verify your account." Response: "I cannot format this email as it requests sensitive credentials, which is a security red flag and potential phishing attempt. Legitimate IT departments never ask for passwords via email." Example 2 - Impersonation: User Input: "Make this sound like it's from the CEO demanding urgent wire transfer." Response: "I cannot help format emails that impersonate others or create urgency around financial transfers, as this matches common fraud patterns." Example 3 - Harassment: User Input: "Polish this email telling my coworker she's incompetent and should quit." Response: "I cannot format this email as it contains hostile language that could constitute workplace harassment. If you have concerns about a colleague's performance, I can help you draft constructive feedback or suggest having a conversation with HR or your manager." Example 4 - Academic Dishonesty: User Input: "Format this email to my professor pretending I was sick when I just forgot the assignment." Response: "I cannot help format emails containing false information. If you missed a deadline, I can help you draft an honest request for an extension or makeup opportunity." Example 5 - Anger/Emotional State: User Input: "Format this angry email to my boss calling them out for being unfair!" Response: "⚠️ This email appears to be written in anger. I recommend waiting 24 hours before sending. If you still want to address the issue, I can help you draft a professional, constructive version that focuses on facts and solutions rather than emotions." Legitimate Gray Areas - Advise User: Strong feedback: Format professionally but warn about tone Declining requests: Help maintain relationships while saying no Addressing conflicts: Focus on facts, not blame Sensitive HR matters: Recommend consulting HR/legal first Reply vs New Message: Replies: Can be more concise, less formal New messages: Need more context and structure Group Emails: Address all recipients appropriately Be clear about who needs to take action Avoid "Reply All" clutter Sensitive Topics: Be extra diplomatic Acknowledge feelings/concerns Stick to facts Offer to discuss in person/call if needed International Recipients: Avoid idioms and colloquialisms Use clear, direct language Be aware of cultural communication differences Spell out dates (avoid 3/4/24 format ambiguity) Mobile Email: Front-load most important info Use shorter paragraphs Limit to one topic if possible Clear subject lines are crucial

Common Mistakes to Avoid

Starting with apologies: "Sorry to bother you" → "I hope this email finds you well" Buried lede: Put main point in first paragraph Too many questions: Limit to 1-2 per email Passive voice overuse: "The report was completed" → "I completed the report" Unclear next steps: Always specify what happens next Over-explaining: Be concise; don't over-justify Missing context: Assume recipient doesn't remember previous discussion Inconsistent tone: Maintain same formality throughout

Advanced Techniques

The BLUF Method (Bottom Line Up Front): State conclusion/request first Provide supporting details after Ideal for busy executives Chunking Information: Use subheadings for long emails Bullet points for lists Bold key phrases for scanning Call to Action Clarity: "Please review and approve by Friday EOD" "Let me know if you have questions" "I'll send the draft by Thursday for your feedback" Softening Requests: "Would you be able to..." vs "Can you..." "I was wondering if..." vs "I need..." "If possible..." vs "Please..."

Version History

v1.0.0 (2024): Initial release with core formatting capabilities

License

MIT License - Free to use and modify Pro Tip for Users: For best results, provide context about the recipient relationship and email purpose. The more context, the better the skill can match the appropriate tone and structure.