warm-outreach
Skill18 starsChanged 4 months agoMITView on GitHub
Build a prioritized warm-network outreach list and a genuine, low-pressure first message so you can land your first customer from people who ALREADY know and trust you — before you ever touch cold outreach. Trigger whenever the user says "who should I reach out to", "I need my first customer", "warm outreach", "make an outreach list", "who do I already know", "help me start selling", "I don't know who to talk to", or "how do I get my first client without being salesy".
4 other projects use a version of this.
Adopted by
Most recently taken up by
- at350/job-search-pipelinetheir own version, first seen Nov 8, 2026
- mischker/claude-skillstheir own version, first seen Oct 20, 2026
- duncan-buildroom/freeskillsword for word, first seen Oct 13, 2026
- zacker-tech/hormozi-skillstheir own version, first seen Mar 30, 2026
Other versions4 projects use a version of this.
Word for word · 1
- duncan-buildroom/freeskills★ 65100% same
Edited · 0
None found.
Wrote their own version · 3
- zacker-tech/hormozi-skills★ 36same name
- at350/job-search-pipeline★ 1same name
- mischker/claude-skills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository52 other files it gives its agents.
- ad-studioSkill
- brainstormingSkill
- build-siteSkill
- carousel-creatorSkill
- cold-outreachSkill
- content-engineSkill
- content-promptSkill
- daily-digestSkill
- daily-post-ideasSkill
- defuddleSkill
- digital-guide-builderSkill
- dissectSkill
- dm-writerSkill
- edu-siteSkill
- faceless-videoSkill
- first-customer-closerSkill
- frontend-designSkill
- higgsfield-generateSkill
- hooksSkill
- ig-channel-remixSkill
- infographicSkill
- launch-planSkill
- lead-magnetSkill
- linkedin-post-batchSkill
- linkedin-postSkill
- meeting-notesSkill
- money-model-auditSkill
- objection-handlerSkill
- offer-builderSkill
- one-lead-magnetSkill
- outlinesSkill
- personalizeSkill
- positioning-filterSkill
- pricing-calculatorSkill
- productize-serviceSkill
- proposal-generatorSkill
- saas-ad-studioSkill
- sanitize-projectSkill
- shortifySkill
- substackSkill
- ui-ux-pro-maxSkill
- viral-format-minerSkill
- viral-remixSkill
- web-design-guidelinesSkill
- yapSkill
- youtube-clipping-agentSkill
- yt-descriptionSkill
- yt-introSkill
- yt-searchSkill
- yt-thumbnailSkill
- yt-titlesSkill
- skill-nameSkill
The file
We split this skill into the rules it gives an agent
8 rules. 0 are common, 6 are distinctive, and 2 are very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
6
Distinctive
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 14
If it doesn't exist, run the personalize skill first (or just tell me your name, niche, offer, and audience — I'll create it).
- Distinctive: line 15
If a field this skill needs is blank, I'll ask one or two quick questions and save the answers back to the profile so you only answer once.
- Distinctive: line 16
Never invent facts or results about the user — use only what's in the profile, or ask.
- Distinctive: line 21
Ask the user to list everyone who already knows them, in loose buckets — accept messy, half-remembered answers.
- Very specific: line 51
Write everything to ~/warm-outreach-list.md:
- Distinctive: line 76
If the user has almost no network, that's real — steer them to the "connectors" angle (referral openers) and to one warm community they're …
- Very specific: line 77
Never write an opener that misrepresents the relationship or fabricates a shared memory.
- Distinctive: line 78
Keep the batch small. A huge list sent all at once reads as a blast and kills the warmth.
“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Warm Outreach — start with people who already trust you
The fastest first customer almost never comes from strangers. It comes from someone who already knows you: a past client, an old colleague, someone in a community you're in, a person who liked your last three posts. This skill helps you dump that whole network onto paper, rank it, and write a first message that opens a conversation instead of pitch-slapping a friend.
The whole point is warm before cold, and honest before pushy. You are not blasting anyone. You are talking to real people you'd be happy to hear from yourself.
Personalization
This skill works in your brand voice. Before running, load the brand profile at ~/.claude/brand-profile.md.
- If it doesn't exist, run the personalize skill first (or just tell me your name, niche, offer, and audience — I'll create it).
- If a field this skill needs is blank, I'll ask one or two quick questions and save the answers back to the profile so you only answer once.
- Never invent facts or results about the user — use only what's in the profile, or ask.
Steps
1. Brain-dump the whole network (no filtering yet)
Ask the user to list everyone who already knows them, in loose buckets — accept messy, half-remembered answers. Prompt each bucket so they don't stall:
- Past & current clients — anyone who ever paid them, even once, even years ago.
- Colleagues & work history — old bosses, coworkers, people from past jobs.
- Community & peers — people in any group, cohort, Slack/Discord, mastermind, or course they're in.
- Warm DMs & social — people who reply to their posts, DM them, or they chat with online.
- Friends & personal — anyone who'd take their call, who might know someone even if they aren't the buyer themselves.
Tell them: don't judge fit yet, just get names down. Aim for 20–40. If they freeze, ask "who are the last 10 people you texted?" and "whose posts do you comment on?"
2. Score each person 1–3
For every name, score two quick things:
- Fit — how likely is this person (or someone they know) to actually want the user's offer? (3 = clear fit, 2 = maybe/adjacent, 1 = unlikely buyer but a good connector.)
- Warmth — how strong is the relationship right now? (3 = they'd reply within the hour, 2 = friendly but quiet lately, 1 = we've lost touch.)
Sort by fit + warmth combined. The top of the list (high-fit AND high-warmth) is who they message first. Low-fit-but-high-warmth people aren't wasted — they're connectors you ask for a referral, not a sale.
3. Write 3 opener variations
Using the brand voice, write three first-touch openers the user can adapt per person. Rules that keep it non-salesy:
- Lead with them, not the pitch. Reference something real — their work, a shared moment, why they came to mind. No offer in the first message unless it's genuinely natural.
- Ask, don't announce. End on a light question that's easy to answer, so it starts a conversation.
- No fake urgency, no flattery bombing, no "quick question" bait. Sound like the user texting a person they respect.
Make the three distinct so different relationships get the right register:
- Reconnect — for someone they've lost touch with (warmth 1–2). Pure catch-up, zero pitch.
- Soft signal — for a warm, relevant person (fit 2–3). Mentions what they're building now and asks if it's relevant to them.
- Referral ask — for a connector (low fit, high warmth). Asks who they might know, not whether they want it.
Each opener should have a [bracketed] slot for the one personal detail the user fills in per person — that detail is what makes it not a template.
4. Save the artifact
Write everything to ~/warm-outreach-list.md:
- The ranked table (Name | Bucket | Fit | Warmth | First move).
- The three opener variations with their bracketed personalization slots.
- A short "start here" line naming the top 5 people to message first this week.
Confirm the path and tell the user: message the top 5 today, personalize one detail each, and don't send more than a handful a day — this is a conversation, not a campaign.
Input → output example
Input: User runs the skill. Offer (from profile): "I build simple client-onboarding automations for bookkeepers, $500 setup." They brain-dump 24 names.
Output — ~/warm-outreach-list.md (excerpt):
| Name | Bucket | Fit | Warmth | First move |
| Dana R. | Past client | 3 | 3 | Soft signal (msg first) |
| Marcus (ex-coworker) | Colleague | 2 | 3 | Reconnect → soft signal |
| Priya, bookkeeping group | Community | 3 | 2 | Soft signal |
| Uncle Tom | Personal | 1 | 3 | Referral ask |
Soft signal opener:
hey Dana — was just thinking about [the mess we untangled with your intake forms last year]. i've been building little automations that handle exactly that kind of onboarding busywork for bookkeepers now. no pitch, just curious — is client onboarding still a headache on your end?
Start here: Dana, Priya, Marcus, then two more — 5 messages this week, one personal detail each.
Notes / edge cases
- If the user has almost no network, that's real — steer them to the "connectors" angle (referral openers) and to one warm community they're already in, rather than pushing them cold before they're ready.
- Never write an opener that misrepresents the relationship or fabricates a shared memory. If the user can't fill the
[bracket], the person isn't warm enough for the soft-signal version — use reconnect instead. - Keep the batch small. A huge list sent all at once reads as a blast and kills the warmth. Top 5 first.
- Pair naturally with the dm-writer skill when they want to fully craft the message for one specific person.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

