5050-gtm
A growing collection of GTM skills for Claude Code, built by 5050Growth.
Skills you can drop into Claude Code (or run as plain Python) to handle the GTM work that doesn't justify a SaaS subscription. CRM-aware, CSV-friendly, no dashboards, no cron jobs, MIT-licensed.
Built for founders, RevOps folks, and Attio consultants who'd rather run a script than buy another tool.
What are Skills?
Claude Code Skills are markdown files that give Claude (or Codex, Cursor, Gemini CLI) specialized knowledge for a specific task. Each skill knows what to ask, where to pull data from, and what to spit out. You invoke them with /skill-name.
The skills here are productized versions of work we do at 5050Growth on real client engagements. We open-source them so the next person doesn't have to rebuild the wheel.
Available Skills
Revenue Analytics
| Skill | What it does |
|---|
/cohort-analysis | Full SaaS cohort analysis from your CRM (Attio/Stripe/CSV) joined to revenue (Stripe/Attio/CSV). Outputs a styled Excel workbook + 11 per-section CSVs. |
/ltv-cac | Forward-looking LTV / CAC / payback model. Four formulas side by side (Skok basic, NDR-adjusted, AI-inference-adjusted, Sequoia contribution-margin) plus a sensitivity heatmap and a 36-month synthetic cohort projection. Verdict anchored on Skok's 3:1 rule. |
What you get from /cohort-analysis: every metric a B2B-SaaS investor or founder asks for, in one workbook:
| KPI | What it means | Where it lives |
|---|
| NRR (Net Revenue Retention) | % of M0 MRR each cohort still pays at M1, M3, M6, M12… The headline retention metric. | Sheet section 2, table 3 |
| GRR (Gross Revenue Retention) | Customer-side retention curve, % of each cohort still paying at M_n. | Sheet section 1, table 3 |
| MRR retention curve | Dollar amount each cohort is paying every month since signup. | Sheet section 2, table 1 |
| Net dollar churn | Signed: positive = churn, negative = expansion (so NDR > 100% shows green). | Sheet section 2, table 2 |
| Customer churn rate | % of cohort lost each month, both vs-base (severity) and vs-previous-month (period rate). | Sheet section 1, tables 4–5 |
| MRR churn rate | Same in dollars, signed so expansion is visible. | Sheet section 2, tables 4–5 |
| CAC payback period | Lifetime month each cohort breaks even on its acquisition cost (given CAC + gross margin). | Sheet section 3 |
| Cumulative gross profit | Dollars of GP each cohort has produced, month by month, vs their CAC. | Sheet section 3 |
| Cohort base size | Customers in each signup cohort + their initial MRR (M0). | Header row of every section |
| Lifetime month cutoff | Per-cohort observable horizon, inferred from your data, newer cohorts only show what's been observed. | Header row + summary CSV |
CAC payback is opt-in (provide a cohort,cac_amount CSV). Gross margin defaults to 80%, configurable. Everything else falls out of customers + revenue alone.
What you get from /ltv-cac: forward-looking unit economics from a handful of inputs:
| KPI | What it means | Where it lives |
|---|
| Skok basic LTV | ARPU × GM / churn, the canonical 3:1 rule reference | Verdict sheet |
| NDR-adjusted LTV | ARPU × GM / (churn − expansion), credits net dollar retention | Verdict sheet |
| AI-adjusted LTV | (ARPU × GM − inference) / churn, handles variable inference cost erosion | Verdict sheet |
| Sequoia contribution-margin LTV | Combines NDR + variable-cost adjustments | Verdict sheet |
| CAC payback | Months to recover CAC; basic and AI-adjusted | Verdict sheet |
| LTV/CAC verdict | 🔴 underwater / 🟡 tight / 🟢 healthy / 🟦 under-investing, anchored on Skok 3:1 | Verdict sheet |
| NDR | Monthly + annual-compounded net dollar retention | Verdict sheet |
| Sensitivity heatmap | LTV/CAC across churn (1–10%) × gross margin (50–90%), see what cutting churn or improving GM does | Sensitivity sheet |
| Cohort projection | 100-customer synthetic cohort over 36 months: customers retained, MRR, cumulative gross profit vs. CAC, line chart | Cohort Projection sheet |
Sources cited inline: Skok (Matrix), a16z, Sequoia, Tunguz.
Sales Ops
| Skill | What it does |
|---|
/disco-prep | Pre-discovery-call prep brief from an intro email, referral, or company name. A 3-section scan doc you read in the two minutes before the call: who they are, 8 ranked scope-defining questions, and the one past win to drop as proof. Output to chat, under 350 words. |
/call-coach | B2B discovery call coach with persistent deal memory. Paste any transcript (Granola, Fireflies, Fathom, plain text, file path) and get a scored coaching report: GAPSI framework flags, gaps the rep failed to build, admissions captured, and a complete next-call script. Tracks deals across multiple calls so coaching compounds over time. Built on the GAPSI framework. |
Data / Enrichment
| Skill | What it does |
|---|
/attio-deepline-enrich | Enrich records inside Attio with a real cost gate. Pick the object (companies/people/custom), the size, and the attributes to fill (emails, phones, firmographics, LinkedIn, ICP fields). It pilots one row to learn the true cost, shows you exact credits + dollars, waits for an explicit yes, runs the enrichment, then writes the values straight back into Attio. Reads/writes are free, so you only ever pay for the enrichment itself, after you approve it. Powered by Deepline. |
Social / Content
| Skill | What it does |
|---|
/linkedin-self-improvement-loop | A build-measure-learn loop for your LinkedIn. Ingests your Creator analytics export, keeps a persistent belief model of what drives reach and engagement, reconciles last cycle's bets against the new data, then proposes ONE experiment and drafts briefs for it. Run it on a cadence and it compounds. Advisory: it proposes, a human posts. No API keys. |
SEO / Demand
| Skill | What it does |
|---|
/gsc-lighthouse | One-shot Google Search Console + Lighthouse health check for a verified GSC property. Sitemap status, per-URL index coverage, 28-day search analytics, and PageSpeed Insights on the top 10 pages, all in one report grouped by theme. |
Finance / CFO
| Skill | What it does |
|---|
/cfo-skill | CFO data dashboard for bootstrapped startups. Two modes: CSV templates (works with any stack, HubSpot/Salesforce/Attio + Mercury/Brex/Qonto + Stripe/Chargebee + Ramp/Pleo/Moss) or live API pull from Stripe + Attio + Qonto + Moss. Outputs an Excel workbook with runway, burn, MRR/ARR, customer concentration, AR aging, DSO, vendor spend, departmental burn. Read-only. Not financial, legal, tax, or investment advice. |
Two modes, same output:
# Mode 1: CSV templates (any stack, zero auth)
python skills/cfo-skill/run.py --source csv --csv-dir ./my-data
# Mode 2: live pull from Stripe + Attio + Qonto + Moss
python skills/cfo-skill/run.py --source api --providers stripe,qonto,moss
What /cfo-skill answers, pulled from the right system:
| Question | Best source | Endpoint |
|---|
| MRR / ARR / customer churn | Stripe (subscriptions) | GET /v1/subscriptions |
| Customer concentration / pipeline / lifecycle | Attio | POST /v2/objects/{companies,deals}/records/query |
| Cash balance, runway, burn rate | Qonto | GET /v2/organization + /v2/transactions |
| AR aging / DSO | Stripe or Qonto | /v1/invoices or /v2/client_invoices |
| Spend by category / vendor / department | Moss | /v1/expenses + /v1/suppliers + /v1/departments |
| Recurring SaaS spend (heuristic) | Moss + Qonto | grouped over time |
Excel output: 8 sheets: Summary (with heuristic flags), Cash Flow, Customers, Concentration, AR Aging, Spend Breakdown, Recurring Vendors, Disclaimer.
Frameworks (LTV:CAC, runway, burn multiple, Rule of 40) and bootstrapped case studies (Mailchimp, Zapier, Basecamp, ConvertKit, Zoho) bundled as references, adapted from EveryInc/charlie-cfo-skill (MIT). Read-only by design, never requests write scopes.
Coming soon
We're adding skills as we productize internal work. Star the repo to follow along:
- Sales Ops: proposal generation from call transcripts, lead scoring rules, follow-up sequencing
- Lifecycle / RevOps: Attio lifecycle audit, ownership rules, list hygiene
- SEO / Demand: programmatic page generators, schema markup audits
- Reporting: Evidence/DuckDB scaffolds for any CRM source
If there's something specific you'd want, open an issue.
Install
Option 1: Project-level skill (recommended)
Copy a skill into your project's .claude/skills/ folder:
git clone https://github.com/NachoLafuente/5050-gtm.git
cp -r 5050-gtm/skills/<skill-name> /path/to/your-project/.claude/skills/
The skill becomes available as /<skill-name> in Claude Code from that project.
Option 2: User-level skill (available everywhere)
git clone https://github.com/NachoLafuente/5050-gtm.git
cp -r 5050-gtm/skills/<skill-name> ~/.claude/skills/
Option 3: Standalone Python CLI
You don't need Claude Code. Each skill is also runnable as a plain Python script:
git clone https://github.com/NachoLafuente/5050-gtm.git
cd 5050-gtm
python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
python skills/<skill-name>/run.py --help
Most skills ship with an examples/ fixture so you can try the pipeline before plugging in real credentials.
Option 4: Fork it
The point of MIT-licensing this is that your stack is yours. If you're on HubSpot + Chargebee instead of Attio + Stripe, fork and add a puller. Most skills are <300 lines of Python, easy to extend.
Philosophy
A few rules we follow when building these:
- CSV is the default. Excel is the most popular GTM tool on earth. Don't fight it.
- No cron, no warehouse. One-shot scripts. If you need a refreshing dashboard, buy ChartMogul.
- No SaaS. Your data stays on your laptop. We never POST anywhere.
- Sources are pluggable. Skills that pull data have a
pullers/ folder. Add one for your CRM in ~30 lines.
- Honest about limits. When a source can't do something (e.g. Attio has no native revenue history), the skill tells you before it runs, not after.
Contributing
PRs welcome. The shape of a new skill:
skills/<your-skill>/
SKILL.md ← frontmatter + instructions for Claude
run.py ← entry point (or whatever the skill does)
examples/ ← optional fixture so people can try without API keys
README.md ← optional, if extra docs help
Match the tone: direct, no fluff, honest about what doesn't work.
Inspiration
Learning from the best, Point Nine and other great B2B-SaaS investors who publish their playbooks publicly (Bessemer, a16z, OpenView, Lenny). These skills turn proven frameworks into one-shot scripts you can run on your own data.
Star history
If this is useful, a star helps a lot, both for visibility and for keeping us motivated to ship more.
Who built this
5050Growth, Attio CRM consultancy. Migrations, integrations, GTM systems for VCs and B2B SaaS.
Founder: Nacho Lafuente (nacho@5050growth.com)
If you want this kind of work built for you on your CRM, that's the day job. Otherwise, fork away.
License
MIT, see LICENSE.
