How does the skill handle data privacy and security?

The skill includes a dedicated step for sanitizing sensitive data. Users are encouraged to summarize tasks and remove proprietary information before the consult script sends data to external providers.

Where are the LLM provider API keys stored?

API keys are stored securely in your local settings.json file under the llm_providers configuration key, manageable via the Command Palette.

Which LLM providers are supported by this skill?

The Multi-LLM Consult skill currently supports Gemini, OpenAI (Codex), and Qwen. You can configure the API keys for these providers within your local settings.

Can I use external models to review my code changes?

Yes, you can use the 'review' purpose to send a patch or context file to an external model for an independent analysis of your code changes.

Are the responses from external LLMs automatically applied?

No. External responses are treated as advisory. Claude will summarize the external feedback into bullet points for your review before any actions are taken.

Multi-LLM Consult

Name: Multi-LLM Consult
Author: NickCrew

byNickCrew

•

데이터 과학 및 ML

Consults external AI models like Gemini and OpenAI to provide second opinions, code reviews, and alternative implementation plans.

소개

The Multi-LLM Consult skill enables Claude to interact with external Large Language Models, including Gemini, OpenAI/Codex, and Qwen, directly from the development environment. It is designed for scenarios where a single perspective isn't enough, allowing users to cross-reference architectural plans, delegate code reviews to specific models, or seek a 'second opinion' on complex refactors. By providing a structured workflow for multi-model comparison and data sanitization, this skill helps developers reduce bias and catch potential logic gaps through independent AI verification.

주요 기능

Integrated support for Gemini, OpenAI/Codex, and Qwen providers
Data sanitization workflow to protect sensitive information before external calls
Standardized output reconciliation to compare and contrast model responses
Support for file-based prompts and context-aware diff analysis
7 GitHub stars
Pre-defined consultation purposes for reviews, planning, and delegation

사용 사례

Comparing alternative logic paths between different LLMs to resolve implementation conflicts
Delegating a tedious documentation or code review task to an external model
Validating a complex refactoring strategy by requesting a second opinion from Gemini

소개

주요 기능

Integrated support for Gemini, OpenAI/Codex, and Qwen providers
Data sanitization workflow to protect sensitive information before external calls
Standardized output reconciliation to compare and contrast model responses
Support for file-based prompts and context-aware diff analysis
7 GitHub stars
Pre-defined consultation purposes for reviews, planning, and delegation

사용 사례

Comparing alternative logic paths between different LLMs to resolve implementation conflicts
Delegating a tedious documentation or code review task to an external model
Validating a complex refactoring strategy by requesting a second opinion from Gemini