During an agent.run stage the task Log tab shows nothing, so the operator can't tell whether the agent is working. On task 462 all nine plan iterations were logged within 2 ms at the end of the stage, after 100 s of silence.
LoopRunner (internal/agentexec/modelloop/loop_runner.go) calls agentloop.Run synchronously; per-iteration lines appear only once it returns.
Done when each model iteration and tool call shows in the task log as it happens.
During an `agent.run` stage the task Log tab shows nothing, so the operator can't tell whether the agent is working. On task 462 all nine `plan` iterations were logged within 2 ms at the end of the stage, after 100 s of silence.
`LoopRunner` (`internal/agentexec/modelloop/loop_runner.go`) calls `agentloop.Run` synchronously; per-iteration lines appear only once it returns.
Done when each model iteration and tool call shows in the task log as it happens.
sam
added the bug label 2026-11-12 15:52:53 +00:00
Fixed in 97fbc35e: LoopRunner now logs every model call and tool call to the task log as it finishes (internal/agentexec/modelloop/progress.go), instead of agentloop's after-the-run step lines being the only trace. task check passed.
Fixed in 97fbc35e: LoopRunner now logs every model call and tool call to the task log as it finishes (internal/agentexec/modelloop/progress.go), instead of agentloop's after-the-run step lines being the only trace. task check passed.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
During an
agent.runstage the task Log tab shows nothing, so the operator can't tell whether the agent is working. On task 462 all nineplaniterations were logged within 2 ms at the end of the stage, after 100 s of silence.LoopRunner(internal/agentexec/modelloop/loop_runner.go) callsagentloop.Runsynchronously; per-iteration lines appear only once it returns.Done when each model iteration and tool call shows in the task log as it happens.
Fixed in
97fbc35e: LoopRunner now logs every model call and tool call to the task log as it finishes (internal/agentexec/modelloop/progress.go), instead of agentloop's after-the-run step lines being the only trace. task check passed.