segmentation
Skill1 starsChanged 56 days agoMITView on GitHub
Use when stage 12 of the prospecting system needs prioritized leads grouped by use case so each group gets its own angle and copy — segments proposed from the customer patterns, approved by a human, then assigned to every account in one classifier pass. Also use when the ask is \"split this list into segments\".
- Reads credentials
4 other projects use a version of this.
Adopted by
Most recently taken up by
- Ken-Technology/cold-email-skillstheir own version, first seen Oct 27, 2026
- Adarshukumar/skillstheir own version, first seen Sep 22, 2026
- 808enzo/chappietheir own version, first seen Sep 17, 2026
- Usermuse/em-plugintheir own version, first seen Jul 31, 2026
Other versions4 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 4
- Ken-Technology/cold-email-skills★ 4same name
- 808enzo/chappie★ 3same name
- Adarshukumar/skills★ 1same name
- Usermuse/em-plugin★ 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 repository21 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- account-notesSkill
- audience-discoverySkill
- audience-filterSkill
- buying-signalsSkill
- cold-emailSkill
- contact-enrichmentSkill
- customer-patternsSkill
- follower-scrapeSkill
- follow-upsSkill
- icp-definitionSkill
- lead-prioritizationSkill
- lead-qualificationSkill
- lead-scoringSkill
- linkedin-outreachSkill
- lookalike-researchSkill
- personalizationSkill
- pipeline-reviewSkill
- prospecting-systemSkill
- reply-classificationSkill
- targeted-searchSkill
The file
We split this skill into the rules it gives an agent
1 rule. 0 are common, 1 is distinctive, and 0 are very specific to this project.
0
Common
Most of the 4 edited versions we read keep these.
1
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 34
Check 10 rows per segment by hand: before joining. If a segment is wrong more than it is right, fix the definition and re-run.
“Common” is measured: kept by at least 6 in 10 of the 4 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
Segmentation
One message for the whole list wastes a good list. This stage splits the list by why they would buy, so stages 13–16 write to a use case instead of an industry.
Pipeline contract
Stage 12 of 20. Reads 11-prioritized.csv, 01-customer-patterns.md. Writes 12-segments.md (definitions) and 12-segmented.csv (adds segment, segment_confidence, segment_reason).
Files live in runs/<run>/. Column definitions: /prospecting-system.
Steps
- Propose 2–5 segments from the customers, not from the list. Each best customer bought for a reason; group those reasons. A segment is a use case + the trigger that makes it urgent ("replacing spreadsheets after a funding round"), not a firmographic bucket ("fintech, 50–200").
- Write
12-segments.md: for each segment a short label, a one-paragraph definition, which customer(s) it came from, what qualifies, what looks close but belongs elsewhere. Add apartnerlabel if vendors or agencies that serve your buyers are in the list — stage 15 treats them differently. - Human approves the segments. They drive all copy downstream.
- Classify accounts, not people — dedupe to unique
account_domain, classify once, join back to leads.
One nano pass assigns each company to ONE segment ornpx tsx engine/list-builder/scripts/classify-batch.ts --csv=<accounts.csv> \ --prompt-file=<run>/12-segments-prompt.txt --labels=<label1,label2,...> \ --out=<run>/12-classified.csv --scrapenone, with confidence and reason.--scrapefetches the homepage when the description is missing. Under ~200 accounts, or with no OpenAI key, use Haiku sub-agents in batches of 25 with the same prompt. - Check 10 rows per segment by hand before joining. If a segment is wrong more than it is right, fix the definition and re-run; do not patch rows.
- Report leads per segment and per tier.
Rules
noneis a valid answer. Leads innoneget bucket copy in stage 13 or sit out; never force a fit.- A segment with fewer than ~30 leads is not worth its own copy. Merge it or park it, and say which.
- More than 5 segments means the ICP is too broad; raise it in the stage 20 review rather than writing 8 campaigns.
- Tiering and multi-segment ICPs:
skills/icp-definition/references/icp-matrix-builder.md.
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.