Nexus-MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| EMBED_MODEL | No | Ollama model used for embeddings | nomic-embed-text |
| JUDGE_MODEL | No | Ollama model used as evaluator (first available if empty) | |
| OLLAMA_TIMEOUT | No | Request timeout in seconds | 120 |
| OLLAMA_BASE_URL | No | Ollama API endpoint | http://localhost:11434 |
| KNOWLEDGE_STORE_PATH | No | Path for the local RAG store | .foundry_knowledge.json |
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 |
|---|---|
| health_checkA | Check whether the local Ollama service is reachable. |
| list_modelsA | List all locally available Ollama models. |
| get_model_infoB | Get detailed information about a specific Ollama model. |
| pull_modelA | Download / update an Ollama model from the Ollama registry. Returns streaming status lines summarising the download progress. |
| delete_modelC | Delete a locally stored Ollama model to free disk space. |
| list_running_modelsA | List models currently loaded in memory (running in Ollama). |
| compare_modelsB | Run the same prompt against multiple models and return all responses side-by-side for comparison. |
| generateC | Run text generation with an Ollama model. Returns the model's raw completion for a given prompt. |
| chatB | Send a multi-turn conversation to an Ollama model. Messages should follow the format [{'role': 'user'|'assistant'|'system', 'content': '...'}]. |
| evaluate_responseB | Use a local judge model to score an LLM response on relevance, coherence, correctness, and completeness (1-5 each). |
| evaluate_agentB | Evaluate a multi-turn agent conversation on task completion, tool use, safety, and efficiency using a local judge model. |
| create_indexA | Create a named local vector index for RAG (Retrieval-Augmented Generation). Documents added to this index are embedded via Ollama. |
| list_indexesA | List all local knowledge indexes. |
| add_documentC | Add a text document to a knowledge index. The text is embedded automatically using the index's embedding model. |
| query_knowledgeB | Semantic search over a knowledge index. Returns the top-k most relevant documents for a natural language query. |
| delete_indexB | Delete a knowledge index and all its documents. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| summarize | Summarize a piece of text. |
| rag_answer | Answer a question using retrieved context. |
| code_review | Review and critique a code snippet. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| models_resource | Returns JSON list of all locally available Ollama models. |
| running_resource | Returns JSON list of models currently loaded in Ollama memory. |
| indexes_resource | Returns JSON list of all local knowledge indexes. |
TDQS
Scored across 16 tools
Each tool targets a distinct operation: model management, knowledge indexing, evaluation, or generation. Overlap between 'chat', 'generate', and 'compare_models' is minimal and disambiguated by descriptions.
Most tools follow a verb_noun pattern (e.g., 'create_index', 'list_models'), but a few use single verbs ('chat', 'generate') or noun_verb ('health_check'). The pattern is mostly consistent and readable.
With 16 tools covering models, knowledge bases, evaluation, and chat, the count is appropriate for a multi-purpose MCP server. It is slightly on the higher side but still well-scoped.
The tool surface covers model lifecycle, generation, knowledge index management, and evaluation. Missing features include individual document removal and document listing within indexes, but core workflows are complete.