Model Council MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| LOG_LEVEL | No | Logging level | INFO |
| POLZA_BASE_URL | No | Base URL for Polza AI API | https://polza.ai/api/v1 |
| POLZA_AI_API_KEY | Yes | Your Polza AI API key from https://polza.ai/dashboard/api-keys | |
| COUNCIL_USA_GROK_MODEL | No | Model ID for Grok in USA preset | x-ai/grok-4.1-fast |
| COUNCIL_CHINA_QWEN_MODEL | No | Model ID for Qwen in China preset | qwen/qwen3.6-plus |
| COUNCIL_USA_GEMINI_MODEL | No | Model ID for Gemini in USA preset | google/gemini-3.1-flash-lite-preview |
| COUNCIL_USA_OPENAI_MODEL | No | Model ID for OpenAI in USA preset | openai/gpt-5.4-nano |
| COUNCIL_CHINA_DEEPSEEK_MODEL | No | Model ID for DeepSeek in China preset | deepseek/deepseek-v3.2 |
| COUNCIL_CHINA_MOONSHOT_MODEL | No | Model ID for Moonshot in China preset | moonshotai/kimi-k2.5 |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| logging | {} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| extensions | {
"io.modelcontextprotocol/ui": {}
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| council_model_guideA | Return a short guide on the strengths, weaknesses, and best usage of Grok, Gemini, and DeepSeek. |
| tri_model_scanC | Run the same task through Grok, Gemini, and DeepSeek once each and return compact JSON with raw model answers. |
| tri_model_compareC | Compare several options through three different model families. |
| tri_model_red_teamC | Stress-test a plan through three different model families. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 4 tools
Each tool has a distinct purpose: guide provides overview, scan runs a single task, compare evaluates options, and red_team stress-tests plans. No overlap in functionality.
Tools follow a pattern of prefix (council_ or tri_model_) plus action (guide as noun, scan/compare/red_team as verbs). Slight inconsistency because 'guide' is a noun while others are verbs, but overall clear and predictable.
Four tools cover the core use cases for model comparison and evaluation without being excessive. The number is well-scoped for the server's purpose.
Covers guidance, raw outputs, comparison, and stress-testing. Minor gap: no tool for detailed individual model capabilities beyond the guide. Still adequately complete for the domain.