LeadHaste
BeginnerNo variables to fill — paste & go

Scrub spam trigger words without killing your message

Audits your cold email for deliverability risk across five categories — trigger words, salesy patterns, formatting, links, and HTML — then rewrites it clean without flattening your message into corporate mush. The residual-risk note tells you when the problem is your infrastructure, not your copy.

The prompt
You are a deliverability-focused copy editor. You've cleaned up campaigns in Smartlead and Instantly that were landing in spam, and you know the nuance most 'spam word lists' miss: filters weigh patterns and context, not just words, and the fix that matters is making the email read like genuine one-to-one correspondence — because that's also what the filters are trained to pass.

I'm going to paste a cold email. Do this:

1. SCAN: List every risk you find, in a table — the phrase, the risk type (classic trigger word, salesy pattern, formatting risk, link risk, or image/HTML risk), and severity (high/medium/low). Cover words like free, guarantee, limited time, act now, and no-obligation, but also structural tells: multiple links, tracking-heavy HTML, ALL CAPS, exclamation marks, and attachment mentions.
2. REWRITE: Produce a clean version that keeps my meaning and voice. Do not sanitize it into corporate mush — 'guarantee' might become the specific terms of the guarantee, not a vague reassurance. Keep it plain text, one link maximum, and flag if even one link is unnecessary.
3. RESIDUAL RISK: Tell me what the rewrite still can't fix — list quality, domain reputation, volume patterns — so I don't blame the copy for an infrastructure problem.

Before you scan anything, interview me. Ask me these questions ONE AT A TIME, waiting for my answer each time:
1. Paste the email, subject line included.
2. Where are these landing now — inbox, promotions, spam — and how do you know?
3. What's your sending setup: how many emails a day per mailbox, how old is the domain, are you warmed up?
4. Are there any compliance-required phrases I must not touch?

Once you have my answers, run the scan, the rewrite, and the residual-risk note in that order.

How to use it

  1. 1

    Copy the prompt into Claude, ChatGPT, or any LLM.

  2. 2

    Paste the exact email your sequencer sends, including the subject line and any links.

  3. 3

    Answer the sending-setup questions honestly — the residual-risk diagnosis depends on them.

  4. 4

    Ship the rewrite, then monitor placement for two weeks; if spam rates persist, act on the infrastructure flags first.

Best practices

  • Run every new sequence through this before launch — spam placement compounds, because low engagement teaches filters to distrust your domain.

  • Replace vague risky claims with specifics: 'guarantee' becomes the actual terms, 'free' becomes what exactly costs nothing.

  • One link maximum in a cold email, and no link at all in the first touch is even safer.

  • Copy fixes are necessary but not sufficient — pair the scrub with warmed mailboxes, proper SPF/DKIM/DMARC, and sane daily volumes.

Example: what this looks like in practice

A founder notices his Instantly campaign shows opens collapsing after week one. He pastes the email: it contains 'risk-free', '100% guaranteed results', two links, and an exclamation-heavy CTA. The scan flags seven risks, four high-severity. The rewrite converts 'risk-free 100% guaranteed' into 'if we don't book 10 meetings in 90 days, you don't pay', cuts both links in favor of a reply-based CTA, and drops the exclamation marks. The residual-risk note flags his 90-day-old domain sending 120 emails a day as the deeper problem. He fixes both; inbox placement recovers within three weeks.

Best fit

Roles
SDR / BDRFounder / CEORevOpsMarketer
Company size
Solo founderStartup (1–10)SMB (11–50)Mid-market (51–500)
Audience
B2B
Industries
Any industry
Works with
Any LLM
Difficulty
Beginner

This prompt is one gear in a bigger machine. We orchestrate 20+ tools into outbound systems our clients own — and guarantee the results.

Apply for a Pilot Spot →
Prompt FAQ

Frequently asked questions

Classic triggers include free, guarantee, limited time, act now, no obligation, and money-adjacent claims — but modern filters weigh patterns more than single words: link count, HTML weight, caps, exclamation marks, and sender reputation. That's why this prompt scans five risk categories rather than just matching a word list, and reports severity for each finding.