Skip to main content
Glama

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
JEV_BRIDGE_PATHNoOverride the default path to the bridge.mjs file. By default the server loads ~/.agents/skills/jev-browser-use/bridge.mjs.
JEV_BROWSER_ACTORNoThe actor name to use for browser access, passed to the bridge.
JEV_CHROME_HEADLESSNoSet to '0' to show the Chrome window. Defaults to headless when unset.

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

CapabilityDetails
tools
{}

Tools

Functions exposed to the LLM to take actions

NameDescription
jev_user_tabsA

List existing browser tabs available to claim.

jev_claim_tabB

Claim one existing browser tab for bounded Jev operation.

jev_browser_runA

Run one Jev chunk in background Google Chrome. Default 12 steps, maximum 30, and 45 seconds. Pass url to open a page. Chrome stays open and the same Jev history is kept. Safe keys are Enter, Escape, Tab, Shift+Tab, PageUp, PageDown, Home, and End. If status is step_limit or budget, inspect the screenshot and call again with the same session_id and goal, without url, only when the task is still valid. If status is needs_verification, stop and verify the screenshot yourself. That status is not a pass. Jev does not type.

jev_waitB

Wait until the open Chrome session accessibility text includes every string in includes and none in excludes. Does not start a Jev decision. Requires session_id.

jev_host_typeA

Type text supplied by the host into the open Chrome session. Jev never chooses this text. Pass field to click that text field first. The typed text is not returned. Then resume with jev_browser_run and the same session_id.

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

TDQS

A3.7/5.0

Scored across 5 tools

Disambiguation5/5

Each tool addresses a distinct step in the Jev browser workflow: enumerate tabs, claim a tab, run a chunk, wait for state, and inject host text. No two tools compete for the same responsibility, and the run/wait/type split is clearly delineated.

Naming Consistency3/5

All names share the jev_ prefix and snake_case, but the internal pattern is inconsistent: claim_tab is verb_noun, user_tabs is a noun phrase, browser_run reverses the order, wait is verb-only, and host_type is ambiguous. Still readable and semantically tied to the domain, so moderately coherent.

Tool Count5/5

Five tools is a focused, appropriate size for a specialized Jev browser automation server. Each tool supports a distinct part of the workflow without redundancy.

Completeness4/5

The core loop is covered: list tabs, claim one, run bounded Jev chunks, wait for accessibility text, and type host-supplied text. Minor gaps exist such as no explicit release/stop tool and the resume path after needs_verification is left to the caller, but the main operation can be completed.

Maintenance

ActivityMaintained
ResponsivenessNo issues