drillan/mixseek-plus

Overview

This skill lists LLM models available to MixSeek-Core and fetches up-to-date provider model info via API. It reports provider-specific prefixes, recommended settings for agent roles, and explicit compatibility notes. It surfaces clear errors when API keys are missing or fetches fail.

How this skill works

The skill queries provider APIs (Google Gemini, Anthropic Claude, OpenAI, Grok, and optional mixseek-plus providers) to retrieve model metadata and capabilities. It returns results in multiple formats (mixseek, json, text, csv), marks code_execution compatibility, and reports API/key errors explicitly. You can filter by provider, output format, and verbosity when running the fetch command.

When to use it

• When you need the latest list of models supported by MixSeek-Core
• Before assigning models to team agents (Leader, Member, Evaluator, code_execution)
• To verify agent_type compatibility (e.g., code_execution support)
• When preparing a team config and needing mixseek-style model identifiers
• To debug missing or invalid API keys during setup

Best practices

• Set the required environment variables for each provider before fetching models
• Use --provider to limit queries when network or key scope is constrained
• Prefer --format mixseek for direct team config copy-paste, json for programmatic integration
• Check code_execution compatibility field and use Anthropic models for code execution agents
• Enable verbose mode to see detailed API errors and resolution hints

Example use cases

• Get all providers' models and copy provider:model-name lines into team config
• Fetch only Anthropic models to find code_execution-compatible model IDs
• Export models as JSON for automated validation of agent_type compatibility
• Run with --provider openai --format text to review model descriptions before choosing a Leader model
• Use mixseek-plus providers to select low-latency Groq models for real-time agents