Open-Testing.AI

Open source

What the maker publishes

Access controls

The site states that the platform supports governed identities, space-scoped authorization, JWT authentication, and a full audit log. source

Agentic testing

Its control plane dispatches steered runs for WCAG, personas, exploratory testing, and regression baselines through registered runners. source

Analytics

Run events feed dashboards, lenses, GraphRAG, and an assistant, with an Entity Query DSL for queries. source

BYOL

The site lists BYOL among the platform’s headline attributes. source

Deployment

The platform runs with Docker Compose on one machine and exposes services behind a single gateway. source

Evidence exports

Runs can be exported as JSON, JUnit, PDF, SARIF, OTLP, or a signed ZIP. source

Failure artifacts

The site says failed runs retain their artifacts. source

Framework conversion

Its Universal Translator converts suites across more than 30 frameworks, including Appium, Playwright, Selenium, Cypress, TOSCA, and UFT. source

Integrations

The platform uses LiteLLM for GraphRAG-backed generation, Kafka for entity events, and offers an MCP surface for agent tools. source

Purpose

Open-Testing.AI describes itself as an open-source AI testing platform for modeling applications, generating and running tests, and analyzing results. source

Skills

The Claude Cowork skills suite covers requirements validation, test design and creation, planning, coverage, reporting, management, and framework conversion. source

Test generation

It imports requirements and generates ISO-29119 test assets with behavior-to-requirement traceability. source

Checked against the maker's pages on 9 Oct 2026.