meta-prompt-engineer

To turn ideal prompt descriptions into robust agents, provide your high-level requirements and draft prompts to generate concrete prompt engineering artifacts, CoT workflows, and MECE-aligned optimization plans.

MECE-aligned prompt engineering decomposes agent tasks into mutually exclusive, collectively exhaustive structures, translating evaluation feedback into structured iteration plans for self-optimizing AI agents.

You can generate multiple prompt candidates using the multi-candidate generation mode, which guides agents from problem understanding to producing structured artifacts alongside initial generation and iterative optimization modes.

Yes, you can use evaluation feedback to automate prompt iteration by translating it into CoT workflows and standardized outputs like prompt.md and changelog.md, enforcing safety constraints during optimization.

The best way to structure prompt engineering artifacts for subagents is to output standardized files like prompt.md and changelog.md, integrating optional references and assets while enforcing safety checks.

No dependencies are required to run meta-prompt-engineer workflows; you only need to provide your ideal-state description and any draft prompts to generate the complete optimization workflow for activation.