Research · Advanced
Automated simulated environment skill learning and cataloging
Direct simulated exploration, write executable control routines, and build an indexed skill catalog
Use with BotBento Markdownالعربية
Opens BotBento. Paste the prompt into a bot; nothing is preconfigured.
- When you askReceive target exploration task
- ThenGenerate and refine control code
- ResultRecord verified routines to library
Run open-ended exploration runs within simulated sandbox environments to discover actions, write reusable programmatic control routines, and catalog verified behaviors into a structured reference database for future experiment trials.
Prompt to start with
BotBentoRecommended
When a simulation goal is assigned, analyze current state variables and generate modular control scripts to achieve the objective. Execute the code through the terminal and monitor runtime errors and environment signals. Iteratively patch bugs until task criteria are satisfied. Before saving or updating routines in the skill database, present the validated script and execution trace to the user for approval. Index all verified behaviors with prerequisite conditions and expected outcomes.