There seems to be a race condition in the TUI when rendering very fast and short responses (e.g. a small model like gpt-5-mini responds to a "hello" very quickly).
When this occurs, resizing the terminal can force the re-render of the response which will then appear correctly
Version affected
c077938 on main branch
How To Reproduce
run cagent run agent.yaml with a small model and send it something short like "Hello"
agent.yaml
version: "2"
agents:
root:
model: openai/gpt-5-mini
instruction: you are a helpful assistant
Expectation
Response always renders correctly
OS and Terminal type
Reproduced on MacOS 26 and Ubuntu 25.10 with vscode and ghostty
Screenshots

There seems to be a race condition in the TUI when rendering very fast and short responses (e.g. a small model like gpt-5-mini responds to a "hello" very quickly).
When this occurs, resizing the terminal can force the re-render of the response which will then appear correctly
Version affected
c077938on main branchHow To Reproduce
run
cagent run agent.yamlwith a small model and send it something short like "Hello"agent.yaml
Expectation
Response always renders correctly
OS and Terminal type
Reproduced on MacOS 26 and Ubuntu 25.10 with vscode and ghostty
Screenshots