Welcome to Aucert
Aucert is an AI-native mobile quality engineering platform. Point it at your mobile app, and it automatically generates test scenarios, executes them on emulators, analyzes the results with visual reasoning, and reports bugs with full reproduction steps — no manual test scripts required.
The 5-layer pipeline
Every test run flows through five stages. Click a layer to learn more:
L1
Generation
→
L2
Execution
→
L3
Analysis
→
L4
Decision
→
L5
Reporting
Click a layer to see details
The pipeline is powered by two cross-cutting systems:
- Knowledge Graph — Builds a rich model of your app by ingesting code ASTs, API schemas, UI layouts, historical test results, and product requirements. This context drives intelligent test generation rather than random exploration.
- Device Twin — A predictive model that bridges the gap between emulator behavior and real-device behavior, adjusting confidence scores for device-specific risks.
Get started
| Guide | What you'll do |
|---|---|
| Quickstart | Install the CLI, connect your app, and run your first test in under 5 minutes |
| Installation | Detailed install guide for npm, yarn, and pnpm with system requirements |
| First test walkthrough | Step-by-step guide to understanding your first test results |
Understand the platform
| Guide | What you'll learn |
|---|---|
| Architecture overview | How the 5-layer pipeline processes your app end-to-end |
| Knowledge Graph | How Aucert builds a model of your app for intelligent test generation |
| Device Twin | How emulator results are adjusted for real-device behavior |
Integrate with your workflow
| Guide | What you'll set up |
|---|---|
| CLI reference | Full command documentation with examples |
| Configuration | Customize test generation, execution, and reporting |
| GitHub Actions | Run Aucert tests on every push or pull request |
| CI/CD integration | GitHub Actions, GitLab CI, and Jenkins setup |
How Aucert is different
Traditional mobile testing requires hand-written test scripts that break with every UI change. Aucert takes a fundamentally different approach:
- No test scripts — The Knowledge Graph understands your app's structure, so tests are generated from context, not from brittle selectors
- Visual reasoning — AI models analyze screenshots to determine pass/fail, catching visual regressions that assertion-based tests miss
- Confidence scoring — Every result includes a confidence score, so you know when to trust automated decisions and when to review manually
- Self-updating — When your app changes, the Knowledge Graph re-ingests the new structure and generates updated tests automatically