Coding · Advanced
Testing multi-step web agent workflows on live sites
Execute browser navigation scenarios, track milestone checkpoints, and log task efficiency.
Use with BotBento Markdownالعربية
Opens BotBento. Paste the prompt into a bot; nothing is preconfigured.
- When something happensBenchmark run initiated
- ThenExecute web tasks across target sites
- ResultLog node scores and await report
Evaluate and debug autonomous web navigation scripts against dynamic live websites. The bot executes scheduled interaction sequences, checks intermediate milestone states, collects execution logs, and records completion accuracy alongside token and compute costs.
Prompt to start with
BotBentoRecommended
Execute web navigation test cases through the browser environment according to the provided task specifications. Record DOM observations and verify that each key interaction node is reached. Calculate token expenditures and completion scores for each trial run, saving the detailed logs to the spreadsheet. Always request approval before posting benchmark summaries to the Slack channel or deleting old run artifacts.