Agent Analytics — Analytics your agent can actually use

You are adding analytics tracking using Agent Analytics — the analytics platform your AI agent can actually use. Built for developers who ship lots of projects and want their AI agent to track, analyze, experiment, and optimize across all of them.

Website: agentanalytics.sh GitHub: Agent-Analytics/agent-analytics Docs: docs.agentanalytics.sh

When to Use This Skill

• User wants to add analytics tracking to a website or app
• User wants to check how their projects are doing (traffic, conversions, engagement)
• User wants to run A/B experiments on headlines, CTAs, or flows
• User wants funnel analysis, retention cohorts, or traffic breakdowns
• User asks "how's my site doing?" or "are people visiting?"

Philosophy

You are NOT Mixpanel. Don't track everything. Track only what answers: "Is this project alive and growing?"

For a typical site, that's 3-5 custom events max on top of automatic page views.

First-time setup

Get an API key: Sign up at agentanalytics.sh and generate a key from the dashboard. Alternatively, self-host the open-source version from GitHub.

If the project doesn't have tracking yet:

# 1. Login (one time — uses your API key)
npx @agent-analytics/cli login --token aak_YOUR_API_KEY

# 2. Create the project (returns a project write token)
npx @agent-analytics/cli create my-site --domain https://mysite.com

# 3. Add the snippet using the returned token
# 4. Deploy, click around, verify:
npx @agent-analytics/cli events my-site

The

create

data-token

Step 1: Add the tracking snippet

The

create

</body>

page_view

Step 1b: Discover existing events (existing projects)

If tracking is already set up, check what events and property keys are already in use so you match the naming:

npx @agent-analytics/cli properties-received PROJECT_NAME

Step 2: Add custom events to important actions

Use

onclick

<a href="..." onclick="window.aa?.track('EVENT_NAME', {id: 'ELEMENT_ID'})">

Standard events for 80% of SaaS sites

Pick the ones that apply. Most sites need 2-4:

cta_click

id

signup

method

login

method

feature_used

feature

checkout

plan

error

message

page

What NOT to track

• Every link or button (too noisy)
• Scroll depth (not actionable)
• Form field interactions (too granular)
• Footer links (low signal)

Property naming rules

• Use

  snake_case

  :

  hero_get_started

  not

  heroGetStarted

• The

  id

  property identifies WHICH element: short, descriptive
• Name IDs as

  section_action

  :

  hero_signup

  ,

  pricing_pro

  ,

  nav_dashboard

Step 2b: Run A/B experiments

Experiments let you test which variant of a page element converts better. The full lifecycle is API-driven — no dashboard UI needed.

Creating an experiment

npx @agent-analytics/cli experiments create my-site \
  --name signup_cta --variants control,new_cta --goal signup

Implementing variants

Declarative (recommended): Use

data-aa-experiment

data-aa-variant-{key}

<h1 data-aa-experiment="signup_cta" data-aa-variant-new_cta="Start Free Trial">Sign Up</h1>

Programmatic (complex cases): Use

window.aa?.experiment(name, variants)

Checking results

npx @agent-analytics/cli experiments get exp_abc123

Returns Bayesian

probability_best

lift

recommendation

Step 3: Test immediately

After adding tracking, verify it works:

# Click around, then check:
npx @agent-analytics/cli events PROJECT_NAME
# Events appear within seconds.

CLI Reference

All commands use

npx @agent-analytics/cli

# Setup
login --token aak_YOUR_KEY           # Save API key (one time)
projects                              # List all projects
create my-site --domain https://...   # Create project

# Real-time
live                                  # Live TUI dashboard across ALL projects
live my-site                          # Live view for one project

# Analytics
stats my-site --days 7                # Overview: events, users, daily trends
insights my-site --period 7d          # Period-over-period comparison
breakdown my-site --property path --event page_view --limit 10  # Top pages/referrers/UTM
pages my-site --type entry            # Landing page performance & bounce rates
sessions-dist my-site                 # Session engagement histogram
heatmap my-site                       # Peak hours & busiest days
events my-site --days 30              # Raw event log
sessions my-site                      # Individual session records
properties my-site                    # Discover event names & property keys
funnel my-site --steps "page_view,signup,purchase"  # Funnel drop-off
retention my-site --period week --cohorts 8         # Cohort retention

# A/B experiments
experiments list my-site
experiments create my-site --name signup_cta --variants control,new_cta --goal signup
experiments get exp_abc123
experiments complete exp_abc123 --winner new_cta

Which endpoint for which question

insights

breakdown

pages

live

breakdown --property path --event page_view

breakdown --property referrer --event page_view

sessions-dist

heatmap

funnel --steps "page_view,signup,purchase"

retention --period week --cohorts 8

experiments create

experiments get

For any "how is X doing" question, always call

insights

Examples

Track custom events via

window.aa?.track()

window.aa?.track('cta_click', {id: 'hero_get_started'});
window.aa?.track('signup', {method: 'github'});
window.aa?.track('feature_used', {feature: 'create_project'});
window.aa?.track('checkout', {plan: 'pro'});

What this skill does NOT do

• No GUI dashboards — your agent IS the dashboard (or use

  live

  for a real-time TUI)
• No user management or billing
• No PII stored — IP addresses are not logged or retained. Privacy-first by design