Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/etrebels/claude-code-growth-osnpx agentmods add commands/etrebels/claude-code-growth-os/retention-reportWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/etrebels/claude-code-growth-os/retention-report)<a href="https://agentmods.dev/commands/etrebels/claude-code-growth-os/retention-report"><img src="https://agentmods.dev/badge/commands/etrebels/claude-code-growth-os/retention-report/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/etrebels/claude-code-growth-os/retention-report"><img src="https://agentmods.dev/badge/commands/etrebels/claude-code-growth-os/retention-report.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00026 | $0.00960 |
| Opus 5.5 | $0.00010 | $0.00384 |
| Sonnet 5.5 | $0.00005 | $0.00192 |
| Haiku 4.5 | $0.00003 | $0.00096 |
Grade A, and why
retention-report scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 47 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Retention Report
Generic chassis variant. If a private operating repo (e.g. your own growth-os) defines a command with this name, that version is canonical for that repo's data.
The monthly readout for the right side of the bowtie. /weekly-review reads the direction; this puts a number on it: are you keeping and growing the revenue you've already won, or filling a leaking bucket? It rolls the account book up to the one number the whole motion answers to — net revenue retention.
From ops/customers.md (the account book, with its ARR column) and its git history over the period:
- Set the period and the starting base. Default to the trailing month (or quarter). The cohort is the accounts you held at the start of the period — reconstruct their starting ARR from the book's git history (
git show <commit-as-of-period-start>:ops/customers.md). New logos signed mid-period don't count toward retention. - Compute GRR and NRR — from the numbers, never eyeballed. Name the formula, then compute:
- GRR (gross revenue retention) = (starting ARR − downgrades − churn) ÷ starting ARR. Caps at 100% — it's the floor, what you keep before any expansion.
- NRR (net revenue retention) = (starting ARR + expansion − downgrades − churn) ÷ starting ARR. Above 100% means the base grew without a single new logo.
- If the book doesn't carry
ARRyet, say so and report the count-based picture instead (logos retained / at-risk / churned / expanded) — and add the column so next month computes.
- Churn by reason. Group the accounts lost or downgraded this period by the real reason — value miss, champion exit, budget, product gap, fit. Use the same reasons
churn-savelogs. The reason column is the roadmap for retention; counts without reasons can't be acted on. - Expansion. Which accounts grew, by how much, and off what trigger — the
EXPANSION-SIGNALsexpansion-signalworked. Net expansion is what carries NRR above 100%. - The one read. State it plainly: is the base net-expanding or net-leaking? Below 100% NRR means acquisition spend is filling a leaking bucket (
docs/operating-model.md— the one number). One line. - Next month's two bets. The single biggest retention risk to clear and the single biggest expansion to chase. Write them to the top of
ops/priorities.md.
Lead with the two numbers and the one read. Don't pad it, and don't invent a number the book can't support — an honest count beats a confident guess. If a CRM is your system of record for ARR, reconcile against it per .claude/rules/crm-usage.md — read-only beyond what the daily rituals already persisted.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago Changed · +18 lines cb69844057b9
- 1mo ago First seen · 29 lines · 26 tokens per session scan A 7cd7cc1f6290
retention-report is a command published in the GitHub repository etrebels/claude-code-growth-os (21 stars, last pushed 7d ago), licensed MIT. It adds 26 tokens to every session and 960 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
go-to-market
Build a complete go-to-market strategy based on the requirements found in files.requirements.
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.