Query external LLMs for second opinions on files with model flags.
npx skills add https://github.com/raine/consult-llm --skill consult-raine
Please help me install this Agent Skill. Skill: consult Source: https://github.com/raine/consult-llm/tree/main/skills/consult Command: npx skills add https://github.com/raine/consult-llm --skill consult-raine
Obtain independent critique from an external LLM within your agent workflow to validate reasoning and decisions.
Load the consult-llm skill and invoke it with one or more model selectors against relevant files.
High-intent search queries and answers about installing and using this skill.
To get a second opinion from another LLM, load the consult-llm skill and invoke it with model selectors and relevant file paths. This queries an external model in parallel to validate reasoning and architecture decisions within your workflow.
Yes, you can run multiple models in parallel for cross-model consultation. By passing one or more model selectors when invoking the skill, you compare reasoning and recommendations across different LLMs in a single prompt-session.
The best way to gather context for an architecture review prompt is using context-aware prompting. The consult skill gathers relevant files and context before querying the LLM to ensure informed, independent critique and validation of your decisions.
Yes, the consult skill supports both browser/web mode and background execution. You can flexibly invoke it with one or more model selectors against relevant files, enabling debate-style synthesis and parallel model collaboration without blocking your workflow.
You need cross-model consultation for debugging when you require independent critique from an external LLM to validate reasoning. It helps identify blind spots by comparing recommendations from multiple models in a single prompt-session.
ahostbrQuery configured external model endpoints and CLI models to synthesize responses for a single question.
alexei-ledOrchestrate a four-phase dialectic debate to stress-test ideas against code.
menoncelloOrchestrate subagents, commands, MCP servers, and skills with parallel execution.
tomlupoInvoke external LLM models to review code, architecture, and design plans.
