Cold outreach
Skill veyralabsgroup/veyraskills/skills/marketing-suite/cold-outreach
Research a prospect and generate a personalized cold outreach sequence. First message, two follow-ups, and a break-up message. Based on real signals from their website, not generic templates.From its SKILL.md
npx -y skills add veyralabsgroup/veyraskills --skill cold-outreachAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 4 stars4 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
5.8 KB, ~1.4k tokens by cl100k_base, as published. Nobody here has run it
Cold Outreach
Research a prospect company and generate a 4-message outreach sequence personalized to their specific situation. Not templates - actual personalization based on what you find.
What this does
Three phases:
- Prospect Research - Extract signals from their website, marketing activity, and context
- Pain Signal Identification - Find the specific angle that will resonate
- Sequence Generation - Write 4 messages: opener, follow-up 1, follow-up 2, break-up
How to use
Pass a company URL or LinkedIn URL:
/cold-outreach https://empresa.com
Or with context:
/cold-outreach https://empresa.com - contacto: Ana Lopez, directora de marketing, LinkedIn: linkedin.com/in/analopez
Phase 1 - Prospect Research
Collect the signals that make outreach feel personal:
From their website:
- What they do and who they serve
- What they are clearly investing in (design, content, ads)
- What is clearly missing or underperforming (compared to what you'd expect)
- Any recent changes (new product, rebrand, expansion)
- Tech stack signals (no pixel = cannot run ads; no analytics = flying blind)
From search:
- Recent news, press, blog posts
- Job postings (reveals priorities and pain)
- Social media activity (are they active? what topics?)
- Reviews or testimonials (reveals what customers value)
From LinkedIn (if profile provided):
- Contact's role and responsibilities
- Recent activity / posts
- Time in current role (new role = often looking to prove impact)
- Shared connections or context
Phase 2 - Pain Signal Identification
Pick ONE primary pain angle. Not five angles - one.
The best angle is:
- Specific to them (you saw something on their site / LinkedIn)
- Connected to a business outcome they care about (revenue, leads, time)
- Something the agency can actually help with
Angle selection:
| Finding | Angle |
|---|---|
| No conversion tracking | "You're spending on [channel] with no way to measure return" |
| No Meta Pixel | "Your audience data is building in competitors' accounts, not yours" |
| Inactive social (3+ months) | "Your social presence is sending the wrong signal to new visitors" |
| No case studies | "Prospects who visit your site can't see proof - that's a closing problem" |
| Old website | "Your website is costing you credibility with every new visitor" |
| SEO competitors ranking above them | "X and Y are capturing searches you should be winning" |
| Hiring for marketing | "You're scaling marketing - this is the moment to get the foundation right" |
What to avoid:
- "I loved your website" (nobody believes it)
- "I noticed you could improve your SEO" (too generic)
- "We work with companies like yours" (irrelevant)
- "I wanted to reach out because..." (filler)
Phase 3 - Sequence Generation
Generate 4 messages. Each has a different purpose.
Message 1 - The Opener
Purpose: Get a reply. One ask. One specific observation.
Format:
Subject: [specific, not clickbait - max 7 words]
[Their name],
[One observation from research - something specific to them]
[One sentence connecting it to a business outcome]
[One question or one offer - not both]
[Name]
Rules:
- Under 80 words
- No attachment, no portfolio link in first message
- The observation must be specific enough that they think "they actually looked at my site"
- One ask only: either a question OR a link/offer, never both
Example:
Subject: Your Meta Pixel isn't firing
Ana,
I was looking at Empresa.com before reaching out - your Meta Pixel is installed but not firing conversion events, which means you're building audiences in Meta without knowing which visits actually convert.
Worth a 15-minute call to show you what that's costing?
Carlos
Message 2 - Follow-up 1 (3-4 days after Message 1)
Purpose: Add value. Give something without asking for anything.
Format:
[First name],
[Reference to message 1 - don't apologize for following up]
[One piece of relevant information or insight - specific to their situation]
[Soft ask or no ask at all]
[Name]
Rules:
- Under 60 words
- Give value before asking again
- Can include a link to a relevant article or example (not your portfolio)
Message 3 - Follow-up 2 (5-7 days after Message 2)
Purpose: Create mild urgency or context. Last real attempt.
Format:
[First name],
[One line: what you've noticed since your last message - could be something about their industry, a competitor, or a seasonal angle]
[The ask again, slightly reframed]
[Name]
Rules:
- Under 50 words
- Do not apologize for following up
- Do not say "I know you're busy"
Message 4 - Break-up (7-10 days after Message 3)
Purpose: Close the loop, leave the door open. Often gets the highest reply rate.
Format:
[First name],
Closing this thread since I haven't heard back. No hard feelings - timing is everything.
[One line: what you'd have helped with]
If things change, you know where to find me.
[Name]
Rules:
- Under 40 words after the opening line
- No negativity, no guilt
- The "what you'd have helped with" line should be specific - it is your last pitch
References
references/email-frameworks.md- PAS, before/after, one-liner formatsreferences/personalization-signals.md- what to look for per company type
Output principles
- One specific observation per message - never generic
- Under 80 words for Message 1, shorter for each subsequent
- One ask per message - never a question AND a link
- No attachments in the sequence (unless asked)
- Personalization must be verifiable - only include things actually found in research
- Sequence must feel human, not automated - if it sounds like a mail merge, rewrite it
What ships with it: 2 files
10.6 KB alongside SKILL.md
references/
- email-frameworks.md5.2 KB
- personalization-signals.md5.4 KB
Gives 0 of the 12 instructions most sales audience skills give in ~1.4k tokens
Counted across 401 of the 401 authors here whose files we hold, read 2026-08-07
- Read product marketing context before asking questionsin 21 of 401, across 11 files
- Acknowledge competitor strengths honestlyin 18 of 401, across 7 files
- Start every page with a summaryin 15 of 401, across 4 files
- Use a single, low-friction call to actionin 15 of 401, across 7 files
- Create a single source of truth for each competitorin 14 of 401, across 3 files
- Make each follow-up email add new valuein 11 of 401, across 5 files
- Cut any sentence that does not drive a replyin 10 of 401, across 4 files
- Tie personalization directly to the problemin 10 of 401, across 4 files
- Write paragraph comparisons for each dimensionin 9 of 401, across 3 files
- Link between related competitor pagesin 9 of 401, across 3 files
- Keep subject lines short and lowercasein 9 of 401, across 3 files
- Define ideal customer profile from top customersin 9 of 401, across 3 files
Said here and by no other author read
- research the prospect company and contact
- pick one primary pain angle
- write a four message outreach sequence
- connect observations to business outcomes
- keep subsequent messages shorter than message one
- include one specific observation per message
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.