ai-orchestration-feedback-loop

Multi-AI orchestration coordinates Claude, Codex, and Gemini in a feedback loop across planning, validation, UX review, and security analysis phases. This Skill enforces role assignments, timeouts, and model configurations to produce traceable validation artifacts and implementation handoffs.

Yes. This Skill supports both triple-AI mode (Claude, Codex, Gemini) and dual-AI modes, letting you scale validation workflows based on task complexity and available resources while maintaining phase-driven validation and security checks.

Multi-AI orchestration produces plans, handoff documents, validation reports, integration logs, and security analysis results. All artifacts are traceable and auditable, enabling you to review decisions and implementation steps across each AI's contribution.

Phase-based workflows assign distinct validation roles to each model—planning with Claude, code synthesis with Codex, UX and security review with Gemini. Orchestration enforces timeouts and role separation, synthesizing their outputs into a single auditable validation record.

This Skill requires model and CLI configurations for Claude, Codex, and Gemini, plus timeout and role-assignment parameters. Setup captures these as part of the orchestration template, ensuring consistent, repeatable validation across runs.

Use multi-AI orchestration for mission-critical software features where planning, security analysis, and UX review must remain independent and traceable. Single-model review works for simpler tasks; multi-AI shines when auditability and cross-perspective validation are required.