You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
core: signature-only responses complete successfully with no visible output #46880
V2 treats a provider response containing only empty reasoning metadata as successful user-visible output.
packages/core/src/session/runner/publish-llm-event.ts sets outputStarted = true on reasoning-start, even if no non-empty reasoning delta arrives. If the provider subsequently emits a clean step-finish (stop / raw end_turn), packages/core/src/session/runner/llm.ts records Step.Ended successfully without checking for text, readable reasoning, or a tool call.
The resulting assistant message can contain only:
{
"finish": "stop",
"rawFinish": "end_turn",
"content": [
{ "type": "reasoning", "text": "", "state": { "signature": "..." } }
]
}
The TUI renders no answer and no error, so the operation appears to have silently done nothing.
Expected: a clean finish with no non-empty text/reasoning and no tool call is surfaced as a no-output provider failure or otherwise handled explicitly. A metadata-only reasoning event should not count as visible output.
Actual: the step succeeds and the session becomes idle with no renderable response.
Plugins
No response
OpenCode version
opencode2 v0.0.0-dev-18900; current V2 source af332a2 remains affected.
Steps to reproduce
In a runner test, stream step-start.
Stream reasoning-start with provider metadata containing a signature.
Stream an empty reasoning delta and reasoning-end.
Stream step-finish with normalized stop and raw end_turn.
Read the projected assistant message and session outcome.
Observe a successful assistant step with no visible content or error.
Screenshot and/or share link
No response
Operating System
Ubuntu 24.04.4 LTS, Linux 5.15.0-91-generic, x86_64
This issue might be a duplicate of existing issues. Please check:
v2: empty reasoning-only response is recorded as successful completion #37372: v2: empty reasoning-only response is recorded as successful completion — same root cause (reasoning-only step with no text or tool call recorded as session.execution.succeeded), same V2 event model. Your report adds useful specifics about publish-llm-event.ts setting outputStarted = true on reasoning-start and the exact content shape with a signature-only reasoning block.
Activity
github-actions commentedon Sep 2, 2026
This issue might be a duplicate of existing issues. Please check:
session.execution.succeeded), same V2 event model. Your report adds useful specifics aboutpublish-llm-event.tssettingoutputStarted = trueon reasoning-start and the exact content shape with a signature-only reasoning block.You may also want to check: