consult-llmlisted

# Consult LLM: Query External AI Models This skill queries external LLM providers via the `consult_llm` MCP tool, sending the same prompt to two models in parallel and presenting both responses for comparison. ## Prerequisites API keys must be configured in **Settings > LLM API Keys** before using this skill. Keys take effect on the next session start. - **OpenAI**: Get your API key from https://platform.openai.com/api-keys - **Gemini**: Get your API key from https://aistudio.google.com/apikey ## Model Selection By default, query both providers in parallel using their latest flagship model: | Provider | Default | |----------|---------| | OpenAI | Latest GPT model available via `consult_llm` | | Google | Latest Gemini model available via `consult_llm` | If the user explicitly names a provider or model (e.g., "ask GPT", "consult Gemini"), only call that provider. The `consult_llm` tool resolves to the best available model for each provider. ## How to Invoke When this skill triggers, follow these steps: 1. **Check for explicit model selection.** If the user specifies a model by name, use only that model. Otherwise, use both defaults in parallel. 2. **Identify what to discuss.** The user's message contains both the trigger and the topic. Extract the topic — it may be: - A question about the current code or architecture - A specific file or function to review - A design decision to evaluate - A general programming question 3. **Construct the prompt.** Buil