hello-stephen opened a new pull request, #66330:
URL: https://github.com/apache/doris/pull/66330

   ## What changed
   
   - Keep the automated reviewer on `gpt-5.6-sol`.
   - Temporarily lower `model_reasoning_effort` from `xhigh` to `high`.
   - Keep Litefuse metadata aligned with the runtime setting.
   
   ## Why
   
   Three of the four review accounts have reached their current Codex usage 
limit. The official Codex guidance recommends `high` for reviewer and 
security-focused agents, while noting that higher reasoning effort increases 
response time and token usage. There is no published fixed `xhigh`-to-`high` 
consumption multiplier because usage depends on the task, context, reasoning, 
and tool calls.
   
   This is intentionally a two-line operational change: model, prompts, 
subagent coverage, routing, permissions, status behavior, timeouts, and job 
topology remain unchanged.
   
   ## Validation
   
   - `actionlint`
   - `git diff --check`
   - bash syntax checks for all 20 embedded `run` blocks
   - verified exactly one runtime `model_reasoning_effort=high` and one 
matching Litefuse `--reasoning-effort high`
   - verified no `xhigh` remains in the workflow
   
   ## Rollback
   
   Restore both values to `xhigh` after account capacity recovers and 
post-change review quality has been sampled.
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to