growth — GrowthScope CLI

AI-first growth product. Terminal-native growth analytics — let a growth Agent (Claude, Cursor, your own scripts) analyze your product alongside you, every day. No dashboards. Just CLI.

Part of the GrowthScope platform — Cursor 之于代码 = GrowthScope 之于增长.

---

🎯 The pitch

Most analytics tools are dashboard-first. They assume you want to open a browser and stare at charts. We don't.

The reality:

• AI agents (Claude, Cursor, your scripts) can run shell commands but can't navigate websites
• Developers in flow can type but not pivot to chart UIs
• Solo founders / small teams don't have time to log into 5 dashboards a day

The fix: a CLI + structured JSON output so an Agent (or you, or your cron) can pull growth data anywhere.

Then ask Claude / Cursor:

"How's MusePod's listen_75 conversion this week vs last? If it dropped, what commit might have caused it?"

It runs growth --json funnel ... + git log + correlates → gives you the answer + a hypothesis.

That's a product growth Agent working alongside you.

---

Install

npm install -g @growthscope/cli

Setup

growth init \
  --name myapp \
  --endpoint https://myapp.com/api/growthscope/v1 \
  --token <admin-token>

Multi-target — add as many products as you have:

growth init --name product2 --endpoint ... --token ...
growth init --name product3 --endpoint ... --token ...

growth targets
# ● myapp     https://myapp.com/api/growthscope/v1
#   product2  https://product2.com/api/growthscope/v1
#   product3  https://product3.com/api/growthscope/v1

Daily-use commands

# What apps are tracking?
growth apps

# Quick stats
growth stats myapp

# Real-time event stream
growth events myapp -n 20

# Funnel
growth funnel myapp app_open feature_used purchase
#  1. app_open          1,234  ████████████████████ 100%
#  2. feature_used        456  ███████ 37% (37% from prev)
#  3. purchase             89  █ 7% (20% from prev)

# Retention matrix (cohort by day)
growth retention myapp

# Anomaly findings (after running scan)
growth findings

# Trigger anomaly detection on-demand
growth scan

Agent-friendly mode (the killer feature)

Every command supports --json for structured output — your Agent's bread and butter:

$ growth --json apps
{"target":"myapp","apps":[{"id":"myapp",...}]}

$ growth --json funnel myapp app_open purchase
{"appId":"myapp","totalUsers":1234,"funnel":[...]}

# Pipe to jq / yq / Agent
growth --json findings | jq '.findings[] | select(.severity=="critical")'

Or --quiet for no colors (CI logs):

growth --quiet stats myapp

Use with Claude Code / Cursor

GrowthScope ships 4 Claude Skills that wrap the CLI:

Skill	Trigger
growth-status	"看一下数据" / "how's myapp doing"
growth-funnel-debug	"为什么 X 掉了" / "why did X drop"
growth-init-product	"add HeyCC to growth"
growth-weekly-report	"周报" / "weekly report"

Drop them into your ~/.claude/skills/ and Claude knows when to invoke growth for you. That's product growth Agent on call 24/7.

All commands

Command	What it does
init	Save target to ~/.growthrc
targets	List configured targets
apps	List apps on a target
stats <appId>	Quick stats (last N days)
events <appId>	Recent events
funnel <appId> <steps...>	Compute funnel conversion
retention <appId>	Cohort retention matrix
findings	List anomalies + opportunities
scan	Trigger anomaly detection on-demand
ask <question>	Natural-language Agent query (V1.1)

Run growth <cmd> --help for options.

Config file

~/.growthrc is YAML:

defaultTarget: myapp
targets:
  myapp:
    endpoint: https://myapp.com/api/growthscope/v1
    token: <admin-token>
  product2:
    endpoint: ...
    token: ...

Privacy

• All credentials in ~/.growthrc (0600 file mode)
• No telemetry — the CLI itself doesn't phone home
• Token sent only to your configured endpoints

Sister projects

• 🟢 growthscope — server + Claude Skills (umbrella, the AI Agent runtime)
• 🟢 growthscope-ios — Swift Package
• 🟢 growthscope-web — Web SDK

License

MIT © 2026 zerolong0