prompt-tunning

What problem does it solve?

This skill automates bias-free prompt tuning for agents by executing prompts and collecting dual evaluations to iteratively improve behavior.

Core Features & Use Cases

• Supports iterative tuning cycles using subagents, with self-reports and directive metrics
• Removes biases and produces auditable improvements across multiple prompts (skills, slash commands, tasks)
• Applicable to AI agents in development, QA, and deployment to stabilize command handling and reduce ambiguity

Quick Start

Initiate a tuning session for a target skill by running a baseline prompt and starting the first iteration to begin collecting feedback.