mata-kadalz
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| VISION_LOG_DIR | No | Directory for logs. | <repo>/runtime/vision/logs |
| VISION_MODEL_ID | No | Model identifier used for cache key. | qwen3-vl |
| LLAMA_SERVER_URL | No | URL of llama-server. Auto-detects WSL gateway IP; set explicitly to override. | http://<gateway-ip>:9931 |
| VISION_CACHE_DIR | No | Directory for cache. | <repo>/runtime/vision/cache |
| VISION_MAX_QUEUE | No | Maximum number of queued inference requests. | 4 |
| VISION_IMAGE_ROOTS | No | Restrict readable image directories. JSON array or comma-separated, relative to repo root. Empty means any path. | |
| VISION_RUNTIME_DIR | No | Directory for runtime files. | <repo>/runtime/vision |
| VISION_MAX_IMAGE_SIZE | No | Maximum image size in bytes (20 MB). | 20971520 |
| VISION_TIMEOUT_SECONDS | No | Timeout for inference in seconds. | 180 |
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
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| vision.inspectA | Analyze a local image file with the Qwen3-VL vision model. Returns structured JSON with the vision analysis. |
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
With only one tool, there is no possibility of confusion between tools. The tool's purpose is clearly defined and unique.
The tool name uses a consistent dot-separated namespace (vision.inspect), which is a clear naming convention. Even with a single tool, the pattern is coherent.
The server has a single tool, which is on the low end. While it serves a specific vision inspection purpose, a single tool feels thin and offers limited functionality.
For the stated purpose of vision inspection, the tool provides a complete analysis capability. There are no obvious missing operations within the scope of inspecting a local image.