ReAct: reason, then act on the AI Agents track. ReAct alternates a short reason with one action. The reason says why this tool, not a speech. The action is one call. The next reason must mention the observation.
This lesson assumes you already worked through Breaking a task into steps.
The idea in practice
Constrain the reason to two or three sentences. Allow only one tool call per turn. If the model emits two calls, run the first and feed the result back before the second.
A concrete check
goal = {
'track': 'AI Agents',
'lesson': 'ReAct: reason, then act',
}
checks = [
'input available at decision time',
'score matches the real decision',
'failure case written down',
]
print(goal['lesson'])
for item in checks:
print('-', item)
Run the sketch locally if you have Python. The printout is a reminder of the checks, not a trained model. Replace the strings with the real inputs from your own example before you treat it as a design.
What usually goes wrong
A long chain of thought with no action, or several tool calls hidden in one turn, breaks the trace and the retry story. When this happens, stop adding parameters or tools. Fix the check, the data, or the permission, then run the same example again.
What to write down
- The input you are allowed to use at decision time.
- The output and the score or pass rule.
- One failure you will test on purpose.
- What you will not claim the system can do.
Practice
Rewrite a messy agent transcript into reason / action / observation rows.
Self-check
- Say ReAct: reason, then act in one sentence that mentions an input and an output.
- Name the failure mode in this lesson and the check that would catch it.
Done when: you can explain this lesson without the page open, and you have a written failure case.