Jev MCP Proxy
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| JEV_DEBUG | No | Set to "1" (default) to log request traces to stderr. | 1 |
| TYPESAFE_API_KEY | Yes | Your API key from the TypeSafe AI Console (https://console.typesafe.ai/keys). | |
| TYPESAFE_BASE_URL | No | Custom API endpoint (default: https://api.typesafe.ai/v1/systemone). | https://api.typesafe.ai/v1/systemone |
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 |
|---|---|
| jev_evaluateA | Primary proxy router for TypeSafe AI Jev. Evaluates unstructured or structured state against multiple typed questions (noul, choice, score) in parallel. Returns answers with calibrated probabilities, confidence scores, and token usage. |
| jev_choiceC | Route a single Choice question to Jev. Categorizes or selects one option from a defined rubric, returning the selected choice, confidence score, and full probability distribution. |
| jev_noulB | Route a single Noul (yes/no) question to Jev. Returns the calibrated probability (0.0 to 1.0) that the answer is yes, suitable for confidence gating and binary routing. |
| jev_scoreB | Route a single Score question to Jev. Rates the input along an ordered rubric of at least two descriptive levels, returning a weighted score and confidence. |
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
jev_evaluate is clearly the multi-question router, while jev_choice, jev_noul, and jev_score each target a distinct question type. There is minor potential overlap because a single question could be sent through either jev_evaluate or the specific single-type tool, but the descriptions make the intended use clear.
All tools share the consistent jev_ prefix and lowercase snake_case style, which makes them recognizable as a family. However, jev_evaluate uses a verb while jev_choice, jev_noul, and jev_score use noun-like type names, so the action pattern is not perfectly uniform.
Four tools is a well-scoped surface for this proxy: one aggregate router plus one tool for each supported question type. There is no redundancy or unnecessary bloat.
The tool set fully covers the apparent domain: the three question types (noul, choice, score) are each directly accessible, and jev_evaluate provides the multi-question parallel path. No obvious missing lifecycle or operational tools are needed for a stateless evaluation service.