Agent skill
customer-story-extractor
Turn raw customer evidence (support threads, QBR notes, call transcripts, Slack messages, NPS comments) into approved, checkable proof: a named case study, quotable lines, and a proof block for the homepage.
Filed under Onboarding, retention and expansion.
From vcxcvii/pipeline-skills · 10 skill entries · 0 · pushed 2026-09-25
What it does when it runs
Turn raw customer evidence (support threads, QBR notes, call transcripts, Slack messages, NPS comments) into approved, checkable proof: a named case study, quotable lines, and a proof block for the homepage. Use when the user says they have happy customers but no case studies, needs testimonials, or their site claims "trusted by leading companies" without naming any.
Automated analysis of the skill and the 0 files bundled beside it. A skill’s own description is written to be selected by an agent, so it describes the job and not the dependencies.
- Keys and connectors you must supply
- None found.
- Hosts it reaches
- No third-party host appears in the skill or its bundled files.
- Tool permissions it declares
- No
allowed-toolsin the frontmatter. It only issues instructions, so there is nothing to bound. - Actions present in the files
- None. Instructions only.
Install it
View source on GitHub ↗git clone --depth 1 --filter=blob:none --sparse https://github.com/vcxcvii/pipeline-skills.git /tmp/pipeline-skills git -C /tmp/pipeline-skills sparse-checkout set "skills/customer-story-extractor" mkdir -p ~/.claude/skills/customer-story-extractor cp -R "/tmp/pipeline-skills/skills/customer-story-extractor/." ~/.claude/skills/customer-story-extractor/
Picked up without a restart. A project skill of the same name is shadowed by your personal one. For one repository only, swap ~/.claude/skills for .claude/skills. Claude Code docs ↗
Or take the whole library
This repo ships a .claude-plugin manifest, so Claude Code can install all 10 skills at once. Plugin skills are invoked as /<plugin>:<skill>, so they never collide with your own.
/plugin marketplace add vcxcvii/pipeline-skills /plugin
The folder is the same in every client that implements the format — 46 of them — so if yours is not above, only the destination changes.
The skill
Source on GitHub ↗Reproduced in full from vcxcvii/pipeline-skills/blob/d284df833e7dd2f21ff15384461ea6fcd3f454c4/skills/customer-story-extractor/SKILL.md, which is licensed MIT (repository). 435 words, 11 headings.
Customer Story Extractor (B2B SaaS)
Buyers discount every claim they cannot check. This skill converts evidence you already have into evidence a skeptic can verify.
Setup
Ask the user to paste any of: a QBR deck summary, a support thread where something went right, a call transcript, renewal notes, or an NPS comment with the free-text field. Two sources is enough to start.
Then ask the one question that decides everything: has this customer agreed to be named, and if not, who would ask them?
Extract
1. Find the before
The state the customer was in, in their words, with a number if one exists. Look for time spent, headcount, error rates, cycle length, or revenue leaking. If there is no number anywhere in the source, say so and use the sharpest verbatim description instead. Do not estimate one.
2. Find the change
What specifically changed, when, and who noticed. The strongest proof is a change a third party observed: the CFO stopped asking, the support queue emptied, the auditor passed it.
3. Find the after
The number again, measured the same way. If the before and after were measured differently, say that plainly in the story. Buyers who catch a mismatched comparison stop believing everything else on the page.
4. Pull the quotes
Three verbatim lines maximum. Keep the customer's grammar. Cleaned-up quotes read like marketing wrote them, because marketing did.
5. Mark what needs approval
Every name, number, and quote gets a status: cleared, needs approval, or cannot use. Nothing without "cleared" goes on a public page.
Output
- Case study draft: situation, what changed, result, quote. Under 400 words.
- Proof block: one sentence with a name and a number, sized for above the fold on a homepage.
- Quote bank: three quotes with role and company, each marked with approval status.
- The approval email: a short, specific request the user can send to the customer, naming exactly what will be published and where.
- Evidence gaps: what is missing to make this checkable.
Rules
- Never invent a customer, a company, a role, or a number. Ever.
- Never publish an unapproved name. An anonymous case study is weak but honest; an unapproved named one is a customer relationship problem.
- Prefer one named customer with a modest number over five unnamed superlatives.
- Round numbers honestly and state the measurement window.
- No em-dashes.
Related skills
sales-one-pager consumes this proof. landing-page-teardown will tell you
where the proof block belongs. positioning-canvas decides which claim the
proof has to support.
Other skills for the same job
Different authors, same problem. Matched on the words in the skill name, across every library in the catalogue except this one.
- customer-story-writer by zapier · 342
- customer-story-builder by edupegoretti · 0
- customer-research by coreyhaines31 · 53,460
- linkedin-hook-extractor by sergebulaev · 4,230
- linkedin-hook-extractor by sergebulaev · 4,230
- where-your-customer-lives by Varnan-Tech · 672
- voice-extractor by BrianRWagner · 440
- find-customer-social-proof by zapier · 342
Need help setting it up?
This page tells you what customer-story-extractor does and what it needs. Cheetah builds the agent setup it runs inside: data, CRM, sequencing and the guardrails.
Book a call →The directory stays free. There is nothing gated behind this.