feat: generic model selection via ACP session/set_model#150
feat: generic model selection via ACP session/set_model#150osolmaz merged 2 commits intoopenclaw:mainfrom
Conversation
Triage resultHuman attention: Quick readThis PR is directionally right and the targeted feature validation passed. Blocking Codex review findings were fixed locally and pushed; no P0/P1 findings remain for current head The remaining handoff reason is CI state: the stored CI snapshot is green, but it is for older head IntentMake WhyBefore this PR, model selection was tied to Claude-specific Codex reviewLocal review for the current branch head is clear on blocking severity. Handled in this lane:
Current review status:
CI/CDTargeted feature validation passed on the PR branch using:
Additional focused reruns were completed locally after fixes in this lane. Stored CI snapshot:
Workflow approval is not the blocker here. The blocker is that the available CI snapshot does not confirm the latest head. RecommendationEscalate to a human for the final merge decision. Recommended next action:
|
Triage resultHuman attention: Quick read
IntentThis PR is trying to make WhyThe underlying problem is that Codex review
CI/CD
RecommendationEscalate to a human. The implementation looks right-shaped, but the flow is blocked by an ambiguous merge conflict in |
Triage resultHuman attention: Quick read
IntentMake Why
Codex review
CI/CD
RecommendationEscalate to a human to decide whether to require completion of the targeted feature validation before landing, or to make an explicit risk call and proceed without it. |
f1c78cc to
b019e9c
Compare
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
|
Landed via temp rebase onto main. Thanks @ironerumi! |
Summary
--model <id>at session creation callssession/set_modelwhen agent advertises modelsacpx <agent> set model <id>routes throughsession/set_modelinstead ofsession/set_config_optionTest plan
tsc --noEmit)Closes #148
Partially addresses #49
🤖 Generated with Claude Code