ollama-ssh-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| OLLAMA_MODEL | No | Default model. | gemma4:e4b |
| OLLAMA_NUM_CTX | No | Context window sent on every request; match other clients of the same model, since a different value forces a reload. | 16384 |
| OLLAMA_SSH_TARGET | Yes | SSH target for the remote Ollama host, e.g. `user@gpu-box.local`; needs key-based, non-interactive SSH. | |
| OLLAMA_REMOTE_PORT | No | Ollama port on the remote host. | 11434 |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
A local LLM (Ollama, default model gemma4:e4b) running on the user's desktop GPU, reached over SSH. Useful for private/offline work (documents that should not leave the house), cheap bulk extraction or classification, and second opinions. It is a small model: verify its output. The first call after ~5 min idle waits ~45s while the model loads; that is expected, do not retry. Pass format (a JSON schema) when you need structured output. Use conversation to keep multi-turn context server-side for the life of this session.
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 |
|---|---|
| chatA | Send a prompt to the local model and return its reply. Args:
prompt: The user message.
system: Optional system prompt. With |
| modelsB | List models installed on the remote Ollama and which are currently loaded in GPU memory. |
| forget_conversationB | Drop a named conversation's history, or all of them if no name is given. |
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 3 tools
Each tool maps to a clearly distinct action: chat runs inference, models lists installed/loaded models, and forget_conversation clears history. There is no plausible way to confuse one for another, even though chat and forget_conversation both touch conversation state (one appends, one deletes).
Two tools use verb-style names (chat, forget_conversation) while models is a bare noun, and chat is a single generic verb rather than verb_noun. The names remain readable and reasonably predictable, but the convention is mixed rather than uniform.
Three tools is a slim but well-scoped surface for a thin Ollama-over-SSH wrapper, and each tool earns its place: generate, enumerate models, reset context. It sits just below the ideal range but does not feel artificially padded or truncated.
Core chat workflow is covered, but there are notable gaps: no model management (pull/delete), no way to inspect or list existing conversations (only drop them), and no stop/stream control. Agents can work around these via the chat tool alone, but the surface is not full lifecycle coverage for the domain.