01
Create projects and track events
Spin up a project, install the tracker, and start collecting page views and custom events.
For
Claude Code
Codex
OpenClaw
Claude Code
Codex
OpenClaw
Your agent can do more than ship code.
It should know what to fix next.
AgentAnalytics lets your AI agent track traffic, find drop-offs, compare projects, and run experiments so you stop shipping blind.
100k free events/month · 2 projects
$1 per 10k events after upgrade
CLI / npx
MCP
API
AgentAnalytics skill
Weekly brief
You ask your agent
What changed across my projects this week?
Your agent replies
my-saas.dev
Signups up 31%.
HN post
340 visits with a 12% bounce rate.
pricing test
The annual-first variant is leading by 18%.
bottleneck
The biggest drop-off is from pricing to signup on mobile.
finding
Most signup growth came from the HN post, not the experiment.
action
Keep the test running and inspect the mobile pricing step.
Why this matters
One answer combines cross-project changes, bottlenecks, experiment results, and the likely explanation, after your agent already did the obvious follow-up analysis.
What this is
Analytics and experiments your AI can inspect, manage, and act on in chat or autonomously.
Create projects, add tracking, and run experiments via CLI, MCP, or API. The AgentAnalytics skill packages the workflows Claude Code, Codex, OpenClaw, and similar tools use to inspect performance, form hypotheses, and review experiments in chat.
Set up tracking once. From there, your AI agent can inspect what changed, question it, and decide what to test next.
What your AI agent can do with AgentAnalytics:
02
Run and manage experiments
Create variants, measure lift, and check winners from the same analytics surface.
03
Use the AgentAnalytics skill
The AgentAnalytics skill packages setup, inspection, and experiment workflows for Claude Code, Codex, OpenClaw, and similar tools.
Dashboards were built for humans. Your AI agent can inspect the data, form hypotheses, and manage experiments in chat.
Use cases
Jobs your AI agent can own once tracking is live.
This is the work traditional analytics leaves on you. AgentAnalytics gives your AI agent a measurement layer it can query directly, so it can review performance, spot what changed, and tell you what deserves attention next.
Portfolio
Cross-project weekly brief
Your agent checks every project, ranks the changes that matter, and gives you a founder-readable brief instead of a metric dump.
- Compare growth across all your active projects in one pass.
- Surface anomalies, winners, and quiet projects that deserve pruning.
- Point you to the one project worth spending time on this week.
Acquisition
Launch and acquisition readout
Your agent can tell whether HN, Reddit, search, docs, or a launch post drove qualified traffic instead of vanity visits.
- Compare channels by visits, bounce rate, and downstream conversion.
- See which landing pages pulled in people who actually kept moving.
- Separate real signal from noisy spikes that look good for a day.
Conversion
Funnel bottleneck detection
Your agent inspects the path from visit to signup or checkout, then tells you exactly where the drop-off happens.
- Break down funnels by page, source, or device to isolate the leak.
- Catch mobile-specific friction before it hides inside aggregate numbers.
- Turn “conversion is down” into a specific broken step to fix.
Experiments
Experiment operator
Your agent can launch, read, and retire A/B tests from code or chat, then recommend the next test instead of stopping at the result.
- Create variants without moving into a separate testing tool.
- Read lift and significance from the same agent-friendly analytics surface.
- Keep the winner, kill the loser, and queue the next hypothesis.
Industries
Built for the AI-era products getting shipped now.
These are not old analytics verticals with an AI label slapped on top. They are the product shapes where builders already rely on coding agents to ship faster and need the same agents to measure what happened.
AI SaaS
AI SaaS and copilots
Track onboarding, activation, pricing, and retention across the product loops where users decide whether your assistant becomes part of their workflow.
Activation
Onboarding
Retention
Pricing tests
Developer tools
Developer tools and APIs
Measure the flow from docs visit to install to first request, then let your agent connect launches and devrel spikes to actual product usage.
Docs to install
First API call
Devrel traffic
Release tests
Content products
Content, docs, and community products
See which articles, docs pages, and community posts attract people who keep reading, click the CTA, and eventually sign up.
Docs to signup
Scroll depth
CTA placement
Source quality
Commerce flows
Commerce and workflow apps
Inspect the steps from landing to demo request or checkout, compare experiments, and find which flows create repeat usage instead of one-time visits.
Checkout funnels
Demo requests
Experiment lift
Repeat usage
Once tracking is live
These are the questions your AI agent can answer and the next moves it can recommend.
Step 01
Weekly brief
Ask "How are my projects doing?" and your AI agent can pull stats, anomalies, top pages, and what deserves attention this week.
Step 02
Funnel bottleneck
Ask where people drop off on the path to signup. Your AI agent points to the exact step losing users.
Step 03
Experiment result
Create an experiment with declarative HTML variants. AgentAnalytics tracks exposures and results while your AI agent reads lift and significance.
Step 04
Next focus
When there is enough data, your AI agent can summarize the winner and point you to the next page, funnel step, or experiment worth attention.
Choose your setup
Self-host for full control, start free on our cloud, or upgrade when you want the full hosted analytics surface.
Open Source
Self-host
Best for teams that want full control over infra, deployment, and data.
Price
Free forever
Hosting
Your infra
Projects
Self-managed
Events
No hosted metering
Agent/API reads
No hosted metering
Retention
You control retention
CLI / API
Full access
MCP
Not included
Basic tracker
Page views, custom events, identify/set
Advanced tracker
Available when enabled in tracker.js
Analytics access
Core analytics surface: project list, stats, recent events
Experiments
Not included
Hosted Free
Start free
See if agentic analytics across multiple projects is something you'd use.
Price
$0
Hosting
Hosted by Agent Analytics
Projects
2
Events
100k / month
Agent/API reads
500 / month
Retention
90 days
CLI / API
Full access
MCP
Limited
Basic tracker
Page views, custom events, identify/set
Advanced tracker
Not included
Analytics access
Core analytics surface: project list, stats, recent events
Experiments
Not included
Hosted Pro
Pay as you go
When you want the full operating loop.
Price
$1 per 10k events
Hosting
Hosted by Agent Analytics
Projects
Unlimited
Events
No monthly cap
Agent/API reads
No monthly cap
Retention
365 days
CLI / API
Full access
MCP
Full access
Basic tracker
Included
Advanced tracker
Impressions, scroll depth, clicks, forms, errors, performance, web vitals
Analytics access
Query + full analytics surface: funnels, retention, sessions, pages, heatmaps, insights, live
Experiments
Included
Open source or hosted
Start free, or self-host it.
Pick the setup that fits now, then start free on our cloud or self-host the open-source version on your own infra.