Supercharge your
playwright tests with AI
Shiplight AI SDK is a developer first toolkit that integrates AI native end to end testing into your existing test infrastructure. It works with your current code, conventions, and CI pipelines while adding intelligence to how tests execute and stay reliable.
Shiplight AI SDK does not replace your test framework. It extends it.
Tests stay in code, follow your repository structure, and go through normal review workflows. Shiplight adds AI native execution, stabilization, and feedback on top.
How It Works
Connect your existing test repository
Install the AI SDK into your current project. No rewrites required. Your fixtures, helpers, utils, and teammates remain unchanged.
Write or generate test logic in code
Tests can be written by engineers or generated by AI. The SDK can generate setup flows, actions, and assertions that match your codebase structure and conventions. AI test logic remains readable and maintainable.
Run tests with AI native execution
During execution, Shiplight uses AI to understand the application as it runs. Tests stabilize past user behavior and system becomes smart of relying on brittle selectors or static states.
Stabilize and self heal tests
When tests encounter flakiness or unstable behavior, the SDK stabilizes execution automatically. Updated logic is generated only when needed and always surfaced as code changes.
Surface results back to CI and systems
Results are reported directly to CI pipelines and internal tools. Failures, regressions, and updates are structured and machine readable. Teams stay in sync through standard review and deployment workflows.
Core Capabilities
AI native actions and assertions
Express intent in clear, deterministic code while Shiplight handles dynamic UI and async behavior during execution.
Full compatibility with existing suites
Bring your existing test project. The SDK adapts to your structure instead of forcing a new one.
Built for CI and automation
Designed for use in CI, local environments, and automated pipelines without special infrastructure.
Deterministic by design
AI enhances execution and reliability, not test randomness. What runs in CI is always rooted in code.
When to Use the
AI SDK
The AI SDK is built for engineering teams who treat testing as code. It bridges the gap between traditional test stability and AI powered adaptability.
Tests must live alongside application code
You need full control over logic, data, and environments
CI pipelines require programmatic execution and gating
AI agents or low-no-code tools generate or modify tests
Reliability and scale matter more than visual abstraction
Relationship to
MCP Server
Shiplight AI SDK and MCP Server are built on the same AI-native execution engine. Many teams use both together to cover different aspects of their testing strategy.
Use The AI SDK
When you want direct, programmatic control over how tests run.
Use MCP Server
When autonomous testing and feedback loops for AI testing agents.