ccg

To get parallel answers from Codex and Gemini for code review, this Skill orchestrates concurrent advisor runs via `omc ask` commands, collecting artifacts from `.omc/artifacts/ask/` before Claude reconciles them into one unified response.

Cross-validating architecture and UX design decisions across multiple AI models is done by decomposing a request into specialized prompts, running Codex and Gemini in parallel, and synthesizing agreements, conflicts, and rationale into a single action checklist.

Yes, you can use Claude to synthesize conflicting outputs from different AI coding assistants by reading their generated artifacts and applying reconciliation logic that surfaces explicit conflicts, rationale, and agreed guidance.

You need specific artifact paths to consolidate multiple AI advisor responses, as the Skill reads outputs directly from `.omc/artifacts/ask/` to ground the final Claude-based synthesis in the actual advisor results.

Running a tri-model orchestration workflow for backend and design tasks requires executing `/oh-my-claudecode:ccg` with a task description that clearly separates Codex and Gemini responsibilities, allowing Claude to merge their outputs.

You should avoid parallel AI model orchestration for development tasks when a single-perspective response is sufficient, as the overhead of decomposing prompts and reconciling artifacts is designed for cross-validation scenarios where tools disagree.