launch-in-app-experiments — Claude Code command for experiments
launch-in-app-experiments is a Claude Code command that plans in-app tests for activation, engagement, monetization, or retention.
Apache-2.0 ★ 450 updated 2026-04-03
.claude/commands/launch-in-app-experiments.mdWhat launch-in-app-experiments does
Provide a product surface, an experiment goal, the number of arms, and metrics to watch as guardrails. You can also describe the audience and add context or risk notes. The command produces a brief covering hypotheses, variants, and instrumentation, a snapshot of the guardrail and monitoring dashboard, and a readout template with a rollout recommendation.
Its workflow moves from hypothesis and success-metric intake to variant design, targeting, triggers, and messaging. It then checks instrumentation, launches and monitors the test, and prepares a summary of results and follow-up decisions. The process invokes usage-growth-scientist and product-adoption-strategist, along with the in-app-messaging-kit and usage-health-scorecard skills. Events, tracking, and guardrail metrics need to be wired before launch; otherwise the monitoring step lacks those inputs.
What it helps with
- turn experiment hypotheses into success metrics
- specify test variants and audience targeting
- prepare event tracking and guardrail checks for launch
- draft a results readout with a rollout recommendation
Example prompts for the agent
- Plan an onboarding experiment for activation with three variants and churn and latency guardrails:
/product-led-growth:launch-in-app-experiments. - Design a monetization test for the dashboard, targeting users by plan:
/product-led-growth:launch-in-app-experiments. - Prepare an experiment readout that connects hypotheses and metrics to a rollout decision:
/product-led-growth:launch-in-app-experiments.
How to install launch-in-app-experiments
Manually: put the files in place
.claude/commands/launch-in-app-experiments.mdCall it with /launch-in-app-experiments. For all projects use ~/.claude/commands/
Manually: put the files in place
~/.codex/prompts/launch-in-app-experiments.mdCall it with /prompts:launch-in-app-experiments. Codex marks this format deprecated in favour of skills
Manually: put the files in place
.github/prompts/launch-in-app-experiments.prompt.mdManually: put the files in place
.gemini/commands/launch-in-app-experiments.tomlGemini stores commands as TOML: the prompt text goes into the prompt field
Questions about launch-in-app-experiments
How do I install launch-in-app-experiments in Claude Code?
Create .claude/commands/launch-in-app-experiments.md in your project and add the command contents to that file.
What does launch-in-app-experiments do?
It coordinates an in-app test from hypothesis and variant planning through monitoring and results review. Its deliverables include an experiment brief, an instrumentation checklist, and a readout template.
What needs to be ready before launching an experiment?
Events, tracking, and guardrail metrics need to be connected so the command can monitor the test against its success and safety measures.
Source text
launch-in-app-experiments.md · Apache-2.0
--- name: launch-in-app-experiments description: Coordinates in-app experiments targeting activation, expansion, or monetization. usage: /product-led-growth:launch-in-app-experiments --surface dashboard --goal activation --variants 3 --guardrails churn,latency --- # Command: launch-in-app-experiments ## Inputs - **surface** – product area (dashboard, onboarding, billing, collaboration, mobile). - **goal** – activation | engagement | monetization | retention. - **variants** – number of experiment arms (including control). - **guardrails** – comma-separated guardrail metrics to monitor. - **audience** – persona/segment targeting (role, plan, region, cohort). - **notes** – optional context or risk flags. ## Workflow 1. **Hypothesis Intake** – capture hypotheses, prior learnings, and success metrics. 2. **Design Package** – define variant specs, targeting, triggers, and messaging. 3. **Instrumentation & Guardrails** – ensure events, tracking, and guardrails are wired. 4. **Launch & Monitoring** – deploy experiment, monitor metrics, trigger alerts. 5. **Readout Prep** – summarize results, decisions, and follow-up actions. ## Outputs - Experiment brief with hypotheses, variants, and instrumentation checklist. - Guardrail + monitoring dashboard snapshot. - Readout template with decision + rollout recommendation. ## Agent/Skill Invocations - `usage-growth-scientist` – leads experiment design + analysis. - `product-adoption-strategist` – ensures journey alignment + messaging. - `in-app-messaging-kit` skill – generates variant messaging + prompts. - `usage-health-scorecard` skill – monitors guardrails + success metrics. ---
© gtmagents · Apache-2.0 · gtmagents/gtm-agents