jev-in-codex
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| JEV_MODEL | No | Optional model name to use for Jev. Defaults to 'jev-latest'. | jev-latest |
| TYPESAFE_API_KEY | No | TypeSafe API key used for Jev relevance evaluations. If not set, the server runs in local fallback mode without TypeSafe ranking. | |
| JEV_WORKSPACE_ROOT | No | Absolute path to the workspace root. Required when the '--root' command-line argument is not provided. An explicit '--root' takes precedence. |
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": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| jev_labelA | Label id/text JSONL text records using Jev and create a complete id/label JSONL artifact. Sends records to TypeSafe. Use feedback_theme or a custom question and 2–16 label definitions. One label per record. Never overwrites files. Returns evidence below confidence 0.8 for review. Benchmark gains apply to the documented workloads; validate accuracy on new taxonomies. |
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 confusing it with another tool. The purpose is singular and well-contained.
With a single tool named jev_label, there are no conflicting naming conventions. The name clearly follows the server's prefix and indicates its labeling purpose.
One tool feels thin for a labeling workflow, but it is not an extreme mismatch since the tool covers the core labeling operation end-to-end. Supporting operations like managing label definitions or reviewing outputs would make the set feel more complete.
The tool covers the main labeling lifecycle: input records, label definitions, artifact creation, and confidence evidence for low-confidence cases. Minor gaps exist around separate review or correction workflows, but the returned evidence mitigates this.