[ Agent API ]

Build once. Let agents manage forever.

The next wave of web traffic isn't human. Use the Humblytics API to let AI agents analyze pages, generate test hypotheses, and launch A/B experiments — programmatically.

split-test-recommendations
GET /api/external/v1/properties/{id}
    /split-test-recommendations

{
  "recommendations": [{
    "title": "Hero CTA Text",
    "selector": "button.hero-cta",
    "confidence": 84,
    "variant": "Start Free Trial",
    "impact": "High: 8%+"
  }]
}

The next wave of web traffic isn't human. AI agents — ChatGPT, Perplexity, Claude — are already crawling, reading, and citing websites. If yours isn't built for them, you're invisible to the fastest-growing source of qualified traffic.

Today, most visitors are people clicking through search results. Tomorrow, most will be agents acting on behalf of people. They don't browse — they read structured data, parse APIs, and make decisions in milliseconds. The websites that win are the ones agents can understand natively.

That's why we built Humblytics API-first. Your agent calls our recommendations endpoint, gets data-driven A/B test ideas with element selectors and confidence scores, and launches experiments — all programmatically. No dashboard. No manual setup. No developer tickets.

Add our endpoints to your CLAUDE.md or .cursorrules and your coding agent becomes a full CRO team. It reads your analytics, spots drop-off patterns, generates hypotheses, and deploys split tests — autonomously. The entire optimization loop, running while you sleep.

The future of software is agentic. The future of websites is agent-readable. The future of optimization is programmatic, API-driven, and always-on. We're building for that future — and making it accessible today.

How it works

Two API calls. Full automation.

Your agent reads your data, generates test ideas, and launches experiments — no human in the loop.

01

Get recommendations

Your agent calls GET /split-test-recommendations?page=/home and receives AI-powered test ideas — each with an element selector, confidence score, current value, and suggested variant. No guesswork.

02

Launch the experiment

Pipe the selectors into POST /split-tests. Your agent creates a no-code A/B test with text changes, CSS modifications, or element swaps — all defined in JSON. The test goes live immediately.

03

Agent manages the rest

Set auto_stop_days: 14 and the experiment runs to completion on its own. Your agent can check back for results, then launch the next test — a continuous optimization loop, fully autonomous.

Use cases

What agents can do with the API

Autonomous CRO agent
Loop through your high-traffic pages, pull recommendations, filter by confidence score, and deploy tests — all without opening a dashboard.
Post-deploy testing
Wire the API into your CI/CD pipeline. Every new landing page automatically gets a split test based on real user behavior data.
Conversion recovery
When your agent detects a conversion drop, it pulls recommendations for the affected page and launches a recovery test targeting the highest-impact elements.
Bulk agency optimization
Manage experiments across dozens of client properties programmatically. Pull recommendations, create tests, and generate weekly reports — at scale.

Get started

Connect your agent in 60 seconds

Add these instructions to your CLAUDE.md, .cursorrules, or system prompt — and your agent becomes a CRO team.

CLAUDE.md
## Humblytics A/B Testing API

API Base: https://app.humblytics.com/api/external/v1
Auth: Bearer $HUMBLYTICS_API_KEY

### Get Recommendations
GET /properties/{property_id}/split-test-recommendations?page={path}

### Create Split Test
POST /properties/{property_id}/split-tests
Body: { name, page, type: "nocode", goal, auto_stop_days,
  variants: [{ label, changes: [{ selector, op, value }] }] }

### Rules
- Always get recommendations first
- Only test recommendations with confidence >= 70
- Default auto_stop_days: 14
- goal: "click_through" for CTAs, "conversion" for forms
The agent web is here

Build for AI agents — or get left behind.

Your competitors are already making their sites agent-readable. Start with the API that lets agents optimize your conversions autonomously.