jev-browser-use-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| TEXT_MODEL | Yes | Text model identifier. If not set, silently defaults to DeepSeek upstream. Validated at startup. | |
| JEV_MCP_BROWSER | No | Set to 'chrome' to enable attached mode, which drives your running Chrome and acts as you in signed-in sessions. | |
| TYPESAFE_API_KEY | Yes | API key for TYPESAFE. Validated at startup. | |
| TEXT_MODEL_API_KEY | Yes | API key for the text model. Validated at startup. | |
| TEXT_MODEL_BASE_URL | Yes | Base URL for the text model API. If not set, silently defaults to DeepSeek upstream. Validated at startup. | |
| TEXT_MODEL_REASONING | No | Reasoning setting for the text model. Example value: 'none'. | |
| JEV_MCP_ALLOW_SCHEMES | No | Comma-separated list of allowed URL schemes. Defaults to 'http,https'. |
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 |
|---|---|
| run_browser_taskA | Drive a browser toward a goal. Start: pass |
| close_browser_sessionB | Close a browser session and its tab. Sessions also expire after 5 minutes idle. |
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 2 tools
The two tools have clearly distinct responsibilities: one drives and continues browser sessions, the other closes them. There is no overlap or ambiguity between starting/continuing work and cleaning up a session.
Both names follow a consistent verb_noun snake_case pattern with the browser prefix: run_browser_task and close_browser_session. The naming is predictable and easy to reason about.
Two tools is on the thin end, but run_browser_task intentionally subsumes a broad range of browser actions through a goal-based interface. It is reasonable for a minimal server, though the surface feels slightly limited.
The core browser-use lifecycle is covered: start a task, continue/redirect it, and close the session. Minor gaps exist, such as no explicit session listing or status tool, but the natural-language task abstraction covers most practical workflows.