second-opinion

Outsource reasoning gaps by triggering an outside-LLM consultation through either repo-aware codex exec or a direct OpenAI API call. This resolves genuine reasoning ambiguity and architecture gaps using structured decision shapes under a shared daily budget.

Use a cross-LLM consultation when your internal analysis hits a genuine gap requiring architectural sanity-checks, texture, experimental prompts, efficiency offload, or signal interpretation. Structured decision shapes guide exactly when to seek outside-LLM advice.

Yes, both the direct OpenAI API and repo-aware codex exec invocation paths share the same daily budget. Budget guards enforce a daily spending cap, consults require mandatory spend disclosure, and you can make optional mid-day budget adjustments.

Provide repo context by routing your request through the codex exec invocation path. This enables repo-aware reads when code context matters, while direct API calls handle general knowledge reads when it does not.

Yes, all outside-LLM consults operate under a shared daily budget with enforced guards and mandatory spend disclosure. This ensures cost discipline and transparency for every architectural sanity-check or reasoning gap consultation.

Outside-LLM consultation resolves genuine reasoning ambiguity that internal analysis cannot. It leverages structured decision shapes to identify when external advice is necessary for signal interpretation, experimental prompts, or texture and real-world nuance.