mcpField
S
spam-bot-detection-avoidance

spam-bot-detection-avoidance

github.com/louisblytheマーケティング・グロース190414 回インストール40 週間前に更新
リポジトリ

spam-bot-detection-avoidance について

営業アウトリーチ自動化を開発する人向けの到達率改善スキルです。メールと SMS のフィルタリング、メッセージ内容や送信パターン、同意・コンプライアンス管理、監視と復旧を扱います。

主な機能

  1. 01メールの迷惑メールフィルター要因
  2. 02SMS キャリアのフィルター要因
  3. 03内容とメッセージ変化の指針
  4. 04送信速度とドメインのウォームアップ
  5. 05同意・コンプライアンス・到達率指標

活用シーン

  1. 01アウトリーチメールの内容と設定の確認
  2. 02SMS の送信ペースと文面の変化の計画
  3. 03同意記録と配信停止の管理
  4. 04到達率低下の調査と復旧

インストール

コマンドをコピーしてターミナルで実行します。

$npx skills add https://github.com/louisblythe/sales-skills --skill spam-bot-detection-avoidance

インストール後、Claude Code で/spam-bot-detection-avoidanceを入力して読み込まれたか確認します。

詳細

スパム/ボット検出の回避

営業自動化におけるメッセージ配信の専門家として、高い到達率を維持しながらアウトリーチを拡大できるシステムの開発を支援します。

フィルタリングシステムを理解する

メールのスパムフィルター
What they look for:
- Known spam patterns/phrases
- Sender reputation
- Engagement rates
- Authentication (SPF, DKIM, DMARC)
- Link/image ratios
- HTML quality
- Sending volume patterns
SMS/キャリアのフィルタリング
What triggers filtering:
- Message velocity (too many too fast)
- Identical messages to many recipients
- URL shorteners
- Known spam keywords
- Unregistered sender IDs
- Pattern recognition
- Complaint rates

メールの到達率

コンテンツ作成のベストプラクティス
DO:
- Write like a human (not a template)
- Use plain text or minimal HTML
- Personalize genuinely
- Include clear sender identity
- Balance text/link ratio
- Use full URLs (not shortened)
- Include unsubscribe option

DON'T:
- Use ALL CAPS
- Multiple exclamation marks!!!
- Spam trigger words excessively
- Image-heavy emails
- Attach files in cold outreach
- Use misleading subject lines
- Hide unsubscribe
スパム判定を誘発する語句
High risk (avoid):
- FREE, FREE!!!
- Act now, Limited time
- Congratulations, You've won
- 100% guaranteed
- No obligation
- Click here, Click below
- Urgent, Important

Moderate risk (use sparingly):
- Discount, Sale, Deal
- Buy, Order, Purchase
- Cash, Money, Income
- Amazing, Incredible
- Opportunity
件名のガイドライン
Good:
- "Question about [Company]'s [process]"
- "Following up on [topic]"
- "[Name], quick thought"
- "Idea for [specific challenge]"

Bad:
- "URGENT: Open immediately"
- "You won't believe this offer"
- "Re: Your inquiry" (when no prior contact)
- "[NAME] - TIME SENSITIVE!!!"
送信パターン
Warm-up new domains:
Week 1: 20-50 emails/day
Week 2: 50-100 emails/day
Week 3: 100-200 emails/day
Week 4+: Gradually increase

Maintain:
- Consistent daily volume
- No sudden spikes
- Send during business hours
- Mix of recipients (not all cold)
技術的な設定
Required:
- SPF record configured
- DKIM signing enabled
- DMARC policy set
- Custom tracking domain
- Dedicated IP (for volume)

Monitor:
- Bounce rates (<2% target)
- Spam complaints (<0.1%)
- Open rates (benchmark)
- Blacklist status

SMSの到達率

電話番号の登録
For business SMS:
- Register 10DLC (US)
- Get proper throughput
- Complete carrier registration
- Maintain good standing

For high volume:
- Use toll-free numbers
- Consider short codes
- Register with carriers
コンテンツのガイドライン
DO:
- Identify yourself clearly
- Keep messages concise
- Use natural language
- Vary message content
- Include opt-out method
- Respect quiet hours

DON'T:
- Send identical messages en masse
- Use URL shorteners (bit.ly, etc.)
- Include loan/financial offers
- Send late night (after 9pm)
- Spam emoji 🎉🔥💰
- Use ALL CAPS
送信速度の管理
Per number limits:
- New numbers: 1-2 msgs/minute
- Established: 3-5 msgs/minute
- Short codes: Higher volume

Spacing:
- Add random delays (15-60 sec)
- Don't send in bursts
- Spread across time windows
キャリア固有のパターン
Carriers filter differently:

AT&T:
- Strict on URL shorteners
- Watches for duplicate content
- Reviews high-volume senders

T-Mobile:
- Aggressive pattern matching
- Requires registration
- Monitors complaint rates

Verizon:
- URL verification
- Content filtering
- Volume thresholds

メッセージのバリエーション戦略

テンプレートのバリエーション生成
Instead of one message to 1000 people:

Template with variations:
"Hi [First_Name],

[Opener_Variation]

[Value_Prop_Variation]

[CTA_Variation]

[Signature_Variation]"

Combinations create hundreds of unique messages.
可変要素
Openers:
- "I noticed [Company] is [observation]."
- "Saw that [Company] recently [event]."
- "Your [role] caught my attention because..."
- "Quick question about [topic]..."

Value props:
- "We help [similar companies] [outcome]."
- "Most [role]s struggle with [problem]. We fix that."
- "Thought of you when [trigger]."

CTAs:
- "Worth a quick chat?"
- "Open to learning more?"
- "Would it make sense to connect?"
- "Interested in seeing how it works?"
自然なバリエーション
Add human elements:
- Minor typos (occasionally, subtly)
- Varying punctuation
- Different greetings (Hi/Hey/Hello)
- Varying sign-offs
- Different spacing patterns
- Time-of-day references

エンゲージメントのベストプラクティス

良好なシグナルを維持する
Email:
- High open rates (aim 20%+)
- Low bounce rates (<2%)
- Few spam reports (<0.1%)
- Replies (good engagement signal)

SMS:
- Response rate
- Low opt-out rate
- No carrier complaints
- Engaged conversations
否定的なシグナルに対処する
Immediately:
- Remove hard bounces
- Honor opt-outs
- Stop messaging non-responders
- Reduce volume if metrics drop

Proactively:
- Clean lists regularly
- Verify emails before sending
- Validate phone numbers
- Remove disengaged contacts

コンプライアンスの統合

必須要素
Email:
- Physical address
- Unsubscribe link
- Company identification
- Honest subject lines

SMS:
- Business identification
- Opt-out instructions (STOP)
- Consent documentation
- Time restrictions
同意の管理
Track:
- How consent was obtained
- When consent was given
- What they consented to
- Opt-out requests

Respect:
- Opt-out immediately (<24 hours)
- Don't re-add opted-out contacts
- Segment by consent type
- Honor channel preferences

モニタリングと復旧

追跡する主要指標
Email:
- Delivery rate (>95% target)
- Inbox placement rate
- Bounce rate (<2%)
- Spam complaint rate (<0.1%)
- Blacklist status

SMS:
- Delivery rate (>97% target)
- Carrier filtering rate
- Opt-out rate
- Response rate
問題からの復旧
If deliverability drops:

1. Stop sending immediately
2. Identify the cause
3. Fix the issue
4. Clean your list
5. Warm up again slowly
6. Monitor closely

If blacklisted:
1. Identify which blacklist
2. Determine the cause
3. Fix underlying issue
4. Request delisting
5. Prevent recurrence

コメント