Config Reference (testronaut.config.js)
{
"provider": "openai",
"model": "gpt-5-mini",
"initialized": true,
"outputDir": "missions/mission_reports",
"projectName": "name_of_your_repo",
"maxTurns": 20,
"sessionToken": "*******************"
}
provider: LLM Provider (openai, gemini, etc...)
model: specific model to use for test runs
reporter: json, html
outputDir: artifact/report location, including screenshots
maxTurns: number of agent actions given per mission to accomplish a goal (minimum 5, maximum 200 - any values beyond these bounds are currently clipped automatically by the CLI to avoid overruns and to better ensure performance)
sessionToken: used for cli comands with authentication needed (i.e. npx testronaut upload)