TraeBridge MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| TRAEBRIDGE_WS_PORT | No | The WebSocket port for the MCP server (default is 8765) | 8765 |
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 |
|---|---|
| browser_navigateC | Navigate the current browser tab (or a new tab) to the given URL. |
| browser_snapshotA | Capture an accessibility-tree (AX) snapshot of the current page. Element handles (@eN) returned by this tool can be used as selectors by the other tools. |
| browser_clickB | Click an element on the page identified by a CSS selector or an @eN snapshot handle. |
| browser_fillB | Fill an input or textarea element (identified by CSS selector or @eN handle) with the given value, replacing any existing content. |
| browser_typeA | Type text into the element that currently has focus, as if the user was typing on the keyboard. |
| browser_send_keysA | Send individual key presses / keyboard shortcuts to the focused element. For typing literal text prefer browser_type. |
| browser_evaluateC | Evaluate arbitrary JavaScript code in the context of the current page and return the result (JSON-serializable values only). |
| browser_screenshotB | Take a screenshot of the current page (or a specific element) and return it as a base64-encoded image. |
| browser_network_startA | Start capturing network requests made by the page. Returns a captureId used with the other browser_network_* tools. |
| browser_network_stopB | Stop a running network capture and return the captured requests. |
| browser_network_listA | List the requests captured so far by an active (or stopped) network capture. |
| browser_network_detailA | Get the full details (headers, body, timing) of one captured request. |
| browser_get_cookiesB | Read cookies from the browser, optionally filtered by URL and/or name. |
| browser_set_cookieC | Create or update a cookie in the browser for the current profile. |
| browser_save_as_pdfA | Print the current page to a PDF and return it as base64-encoded data. |
| browser_uploadB | Attach local files to a file input element on the page (as if picked via the file chooser). |
| browser_list_tabsA | List all open browser tabs with their id, URL, title and active state. |
| browser_switch_tabB | Activate (focus) the given browser tab. |
| browser_close_tabB | Close the given browser tab. |
| browser_cdpC | Send an arbitrary Chrome DevTools Protocol command to the current tab and return the raw response. Use with care. |
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 20 tools
Tools are clearly separated by action and target, with explicit differentiation between browser_fill (replace content), browser_type (literal typing into focus), and browser_send_keys (key presses/shortcuts). Network capture and cookie tools form distinct logical groups. No significant overlap or ambiguity.
All 20 tools use a consistent snake_case pattern with a browser_ prefix, followed by either verb_noun (browser_navigate) or noun_verb (browser_network_start) conventions. There are no mixed camelCase or inconsistent verb styles.
20 tools is slightly above the typical 3–15 range but appropriate for a comprehensive browser automation server. Each tool covers a unique capability (navigation, interaction, network, cookies, tabs, CDP) and earns its place.
The surface covers core browser workflows including navigation, interaction, snapshots, screenshots, network capture, cookies, PDF, upload, and tabs. Minor gaps exist for explicit back/forward/reload, select option, hover, or wait primitives, but browser_evaluate can work around most of them.