Local AI Test Case Generator
Budget: ₹12,500 – ₹37,500 INR
I need an AI-driven desktop application that lets me craft end-to-end test cases through a friendly GUI, then instantly turns those scenarios into ready-to-run JavaScript code. Inside the interface I must be able to choose whether the script is written for Cypress or for Playwright; once I pick the framework, the tool should generate syntactically correct tests, package them as standalone files, and offer a direct download button.
Key points I need met:
• Runs entirely on the user’s machine—no server calls—so the generated code, prompts, and any test data never leave the local environment.
• A clean, intuitive GUI where I describe the flow (page actions, assertions, data inputs, etc.) and watch the AI suggest, refine, and output the final script.
• One-click download of the finished test file in a typical project structure (e.g., spec file plus minimal setup) ready to be dropped into an existing repo.
• The architecture must prevent accidental remote execution; for instance, offline-first packaging, no telemetry, and code signing or hashing to reassure users the binary hasn’t been tampered with.
I’m open to your recommended tech stack for the shell itself (Electron, Tauri, native, etc.), provided the resulting app remains lightweight, secure, and cross-platform. Please outline how you’ll:
1. Handle local AI inference or bundle calls to a model while preserving privacy.
2. Translate the user’s test flow into valid Cypress/Playwright syntax.
3. Implement the GUI elements and download mechanism.
4. Package and distribute the app for Windows, macOS, and (ideally) Linux.
If you have prior examples of similar dev-tool work or AI-assisted code generators, that will help me evaluate fit.
Key points I need met:
• Runs entirely on the user’s machine—no server calls—so the generated code, prompts, and any test data never leave the local environment.
• A clean, intuitive GUI where I describe the flow (page actions, assertions, data inputs, etc.) and watch the AI suggest, refine, and output the final script.
• One-click download of the finished test file in a typical project structure (e.g., spec file plus minimal setup) ready to be dropped into an existing repo.
• The architecture must prevent accidental remote execution; for instance, offline-first packaging, no telemetry, and code signing or hashing to reassure users the binary hasn’t been tampered with.
I’m open to your recommended tech stack for the shell itself (Electron, Tauri, native, etc.), provided the resulting app remains lightweight, secure, and cross-platform. Please outline how you’ll:
1. Handle local AI inference or bundle calls to a model while preserving privacy.
2. Translate the user’s test flow into valid Cypress/Playwright syntax.
3. Implement the GUI elements and download mechanism.
4. Package and distribute the app for Windows, macOS, and (ideally) Linux.
If you have prior examples of similar dev-tool work or AI-assisted code generators, that will help me evaluate fit.