multi-llm-consult

The 'multi-llm-consult' skill enables users to consult multiple external Large Language Models (LLMs) such as Gemini, OpenAI/Codex, and Qwen for alternative perspectives, second opinions, or task delegation. By querying these LLMs with sanitized prompts, this skill provides valuable insights that assist in decision-making and problem-solving processes. It ensures users have access to diverse viewpoints, helping them make informed choices, particularly when comparing answers or evaluating different approaches to a given task.

The skill is structured around a consult script that allows users to specify the purpose of the consultation, whether it's for a second opinion, task review, alternative planning, or delegating subtasks. It helps users summarize their tasks, sanitize sensitive data, and compare responses from different LLM providers. The flexibility to delegate tasks like code reviews or plan checks to these models ensures that users can offload some of their workload. The responses from the LLMs are advisory, and users are encouraged to verify the recommendations against the project's current state.

Target users for this skill include developers, project managers, and anyone involved in decision-making or collaborative tasks who would benefit from getting a diverse set of inputs from various LLM providers. It is particularly useful in complex projects where multiple perspectives are necessary to ensure comprehensive decision-making.