Search: "iterative"

Generates scientific diagram drafts using Nano Banana 2 AI with smart iterative refinement. Uses Gemini 3.7 Flash for quality review. Refines when the review requests improvement, with at most two generations. Specialized in neural network architectures, system diagrams, flowcharts, biological pathways, and complex scientific visualizations.

Runs an autonomous review-and-fix improvement loop over a Claude Code skill until a review comes back clean, with a cross-round findings ledger, escalation when fixes stop converging, and a mechanical scope guard. Reviews are performed by the plugin-dev skill-reviewer agent. Use to fix skill quality issues, iteratively refine a skill, or resume a loop after an escalation ('fix my skill', 'improve this skill until it passes review', 'skill improvement loop'). NOT for a one-time review — use the plugin-dev skill-reviewer agent directly.

An multi-agent design-to-code tool that generates production-ready React code with high visual fidelity and iterative validation.

Attempt at A. G. I. Strive to enhance it through iterative processes, continuously refining and optimizing each version to achieve better outcomes over time (refer to the Readme for additional information that provides further insights and guidance on the Quillan v4.2 project).