teshi-engine, developed by lilinfangrelax, is an execution engine that targets the automated testing lifecycle from capture through repair. The app records interactions, generates PageObjects and Gherkin step definitions, executes tests, and applies automatic healing to broken locators using a protocol-driven pipeline. Key capabilities include an NDJSON automation protocol, a terminal interface, and automated project workspace creation. It targets developers and QA automation engineers focused on reducing test maintenance effort.

What does teshi-engine do?

The tool follows a four-step automation cycle that moves from capture to repair, implementing a unified NDJSON protocol to pass events between components. In practice that means it can record, generate, execute, and heal within a single workspace, producing organized directories for features, recordings, and logs. This workflow is oriented to web and API testing and supports Gherkin-based BDD scenarios to align tests with specifications.

How does teshi-engine affect system resources during runs?

The app exposes a terminal user interface for session control while delegating browser automation to Playwright, so measurable CPU and memory usage mainly reflect the Playwright browser instances rather than a large central engine process. Installation through the Windows package manager and command-line operations that require a Python environment mean setup and runtime activity are local. Heavy runs therefore consume resources from the browser processes launched for web automation.

Is teshi-engine safe to use on production machines?

The tool creates workspace directories and writes generated PageObjects, step definitions, and logs to disk, so it performs file-system writes under normal use. Its automatic healing modifies locators during execution, which affects test artifacts; that behavior suggests keeping test projects under version control and isolating test accounts. Presence in the Windows package repository and active maintenance indicate community acceptance, but run experiments outside production systems first.

Do I need technical knowledge to operate teshi-engine?

The app automates code generation and offers a TUI for recording and playback, which reduces manual coding for automation engineers. Practical use requires familiarity with Playwright-based automation and a local Python environment for certain command-line tasks, so the tool is best suited to developers and QA engineers who can manage framework setup. Non-technical users should expect an onboarding curve before effective use.

Final assessment: suited to engineering-led QA teams

teshi-engine is best suited to engineering-led QA teams that can allocate setup time and stewardship, rather than teams seeking out-of-the-box simplicity. Expect an onboarding period and plan test runs on dedicated agents to reduce local resource contention. For pilot deployments, stage the engine on small projects to validate core workflows before broad rollout. This limits integration risk and validates whether maintenance gains justify the setup effort. Recommended.