🤖 What AI Agents Actually Do Perceive · Reason · Act

🔍 Perception

Click Run or Step to start.

⚙ Rule

Waiting for first step…

🎯 Action

—

Cycle 0 Goals 0 / 2 Steps 0
Core p5 logic too long repetitions – final version below. I understand the issue. Let me produce the polished final complete HTML with proper rendering and obstacle toggling inside p5. I'll ensure no syntax errors, correct BFS, click to toggle obstacles, and animate agent along path.

Perceive, Reason, Act: Inside an AI Agent's Loop

Read the explanation

An autonomous AI agent operates in an endless feedback loop: first sensing its surroundings, then reasoning toward an objective, and finally executing an action. During perception, the agent scans its surroundings. In this simulator's default mode, a three-cell Manhattan radius identifies visible goals and impassable obstacle walls. In the reasoning phase, the agent executes breadth-first search across the grid to discover the shortest unblocked path to the nearest reachable goal. Finally, the agent acts by stepping exactly one cell along the planned route. Clicking Run or Step in the tool lets you watch each cycle re-evaluate its path in real time.

Super generates helpful tools and automates fact-checking across the internet proactively. If you enjoyed this tool, build your own with Super and share it with a friend.