typesafe-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| JEV_BACKEND | No | Force a specific backend: 'typesafe', 'openrouter', or 'vercel'. Optional. | |
| TYPESAFE_API_KEY | No | TypeSafe API key for the default backend. Required if using the TypeSafe backend directly. | |
| TYPESAFE_BASE_URL | No | Override the base URL for the TypeSafe backend. | |
| AI_GATEWAY_API_KEY | No | API key for the Vercel AI Gateway backend. Required if using the AI Gateway. | |
| OPENROUTER_API_KEY | No | API key for the OpenRouter backend. Required if using OpenRouter. | |
| AI_GATEWAY_BASE_URL | No | Override the gateway host for the Vercel AI Gateway backend. | |
| OPENROUTER_BASE_URL | No | Override the host for the OpenRouter backend. | |
| TYPESAFE_API_KEY_ENV | No | Name of an alternative environment variable that holds the TypeSafe API key. | |
| AI_GATEWAY_API_KEY_ENV | No | Name of an alternative environment variable that holds the AI Gateway API key. | |
| OPENROUTER_API_KEY_ENV | No | Name of an alternative environment variable that holds the OpenRouter API key. | |
| TYPESAFE_DEFAULT_MODEL | No | Override the default model for the TypeSafe backend. |
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 | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| evaluateA | Evaluate content against one or more typed questions using TypeSafe's System One API (model 'jev'). Supports three question primitives that can be mixed in a single call: noul (yes/no, returns a probability 0-1), choice (select the best-fitting option from a fixed set), and score (rate against a rubric with 2-10 levels). All questions in a call are answered in parallel against the same input. |
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 1 tool
There is only one tool, so there is no possibility of an agent confusing it with another. The description clearly defines what 'evaluate' does and its supported question primitives.
With a single tool, naming inconsistency is impossible. 'evaluate' is a clear, action-oriented verb that accurately describes the tool's purpose.
One tool feels thin for a server named 'typesafe-mcp', even though it is not trivial because it packages multiple evaluation primitives. The minimal surface is workable but leaves little room for supporting operations.
The tool covers the full stated domain: noul, choice, and score questions can be mixed and answered in parallel. No obvious required operations are missing for this narrow evaluation-focused use case.