Conclave MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| OPENROUTER_API_KEY | Yes | Your OpenRouter API key (get from https://openrouter.ai/keys) |
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": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| conclave_quickA | Query the conclave for quick parallel opinions (Stage 1 only). Fast and cheap - queries all conclave models in parallel and returns their individual responses. No peer ranking or synthesis. Good for getting diverse perspectives quickly. If a custom conclave is active (via conclave_select), it will be used instead of the tier-based config. Args: question: The question to ask the conclave tier: Model tier - "premium" (frontier), "standard" (default), "budget" (cheap/fast) Ignored if custom conclave is active. Returns: Individual responses from each conclave model |
| conclave_rankedA | Query the conclave with peer rankings (Stage 1 + 2). Medium cost - gets individual opinions, then has each model anonymously evaluate and rank all responses. Returns aggregate "street cred" scores showing which models performed best on this specific question. If a custom conclave is active (via conclave_select), it will be used instead of the tier-based config. Args: question: The question to ask the conclave tier: Model tier - "premium" (frontier), "standard" (default), "budget" (cheap/fast) Ignored if custom conclave is active. Returns: Individual responses plus aggregate rankings |
| conclave_fullA | Run the full conclave with synthesis (all 3 stages). Most comprehensive - collects opinions, peer rankings, then has a Chairman model synthesize the best possible answer from the collective wisdom. If a custom conclave is active (via conclave_select), it will be used instead of the tier-based config. The custom chairman overrides the chairman and chairman_preset parameters. Args: question: The question to ask the conclave tier: Model tier - "premium" (complex), "standard" (default), "budget" (simple) Ignored if custom conclave is active. chairman: Override chairman model (e.g., 'anthropic/claude-sonnet-4') Ignored if custom conclave is active. chairman_preset: Use a context-based preset - "code", "creative", "reasoning", "concise", "balanced" Ignored if custom conclave is active. Returns: Chairman's synthesis, consensus level, rankings, and individual responses |
| conclave_configA | View current conclave configuration. Shows conclave member models, current chairman with rotation info, available chairman presets, consensus thresholds, and API key status. Also shows custom conclave selection if active. Returns: Current configuration as formatted JSON |
| conclave_estimateA | Estimate cost for a conclave query before running it. Provides approximate cost breakdown for quick/ranked/full query types. Args: question: The question (used to estimate token count) tier: Which tier to estimate - "quick", "ranked", "full" (default: all) Returns: Cost estimates for each query type |
| conclave_modelsA | List all available models with selection numbers. Shows all models from all tiers with unique numbers that can be used with conclave_select to create a custom conclave. Numbers are stable:
Returns: Numbered list of all available models grouped by tier |
| conclave_selectA | Create a custom conclave from model numbers. Select specific models by their numbers (from conclave_models). The first model in the list becomes the chairman. This custom selection persists until server restart or conclave_reset. Args: models: Comma-separated model numbers, e.g. "1,5,11,14" First number = chairman, rest = conclave members Returns: Confirmation of the new conclave configuration Example: conclave_select(models="31,1,11,21") creates: - Chairman: model #31 (deepseek-r1) - Members: models #1, #11, #21 |
| conclave_resetA | Clear custom conclave selection and return to tier-based config. After reset, queries will use the tier parameter (premium/standard/budget) instead of the custom model selection. Returns: Confirmation that custom selection was cleared |
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 8 tools
Each tool has a clearly distinct purpose with no ambiguity. For example, conclave_quick, conclave_ranked, and conclave_full represent distinct stages of query processing, while conclave_select and conclave_reset manage custom configurations, and conclave_models and conclave_config provide informational views. The descriptions clearly differentiate their roles.
All tool names follow a consistent snake_case pattern with a 'conclave_' prefix and descriptive suffixes (e.g., conclave_config, conclave_estimate, conclave_full). This uniformity makes the tool set predictable and easy to understand, enhancing usability for agents.
With 8 tools, the server is well-scoped for its purpose of managing and querying a conclave of models. Each tool serves a specific function, such as configuration, estimation, querying at different stages, and model selection, without redundancy or unnecessary complexity.
The tool set provides complete coverage for the conclave domain, including configuration viewing, cost estimation, querying at all stages (quick, ranked, full), model listing, custom selection, and resetting. There are no obvious gaps; agents can perform the full lifecycle from setup to querying and cleanup.