chatgpt-remote-browser-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| OPENCLAW_BIN | No | OpenClaw executable; default /opt/homebrew/bin/openclaw | /opt/homebrew/bin/openclaw |
| OPENCLAW_BROWSER_PROFILE | No | Browser profile; default chrome | chrome |
| OPENCLAW_BROWSER_TIMEOUT_MS | No | Backend timeout; default 30000 | 30000 |
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 |
|---|---|
| list_shared_tabsA | List only Chrome tabs currently permitted by the OpenClaw Selected Tabs access boundary. |
| snapshotA | Read the accessible page structure and text from a currently shared tab. |
| screenshotC | Capture an image of a currently shared tab. |
| clickA | Click an element in a currently shared tab. This changes browser or external-site state. |
| typeC | Type text into an element in a currently shared tab. This changes browser or external-site state. |
| pressB | Press a keyboard key in a currently shared tab. This can change browser or external-site state. |
| scrollC | Scroll a currently shared tab by one or more pages. |
| navigateA | Navigate a currently shared tab to an HTTP(S) URL. This changes browser and external-site state. |
| native_browserC | Full high-level OpenClaw Chrome-extension browser surface for Selected Tabs. Use list_shared_tabs first and pass its opaque handle. Supports tab lifecycle, DOM interaction, evaluation, diagnostics, emulation, files and batch actions. Raw shell and raw target IDs are never accepted. |
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 9 tools
native_browser is a full browser surface that overlaps with nearly every specific tool (navigate, click, type, scroll, screenshot, etc.), making tool selection ambiguous. The specific tools are individually distinct, but the presence of an umbrella tool blurs the boundary between simple actions and the high-level API.
Most tools follow a consistent imperative verb pattern in snake_case (navigate, click, type, scroll, press, snapshot, screenshot, list_shared_tabs). native_browser breaks the pattern by being a noun rather than a verb, but the overall convention is still readable and predictable.
Nine tools is a reasonable size for a browser-control server and fits the well-scoped 3-15 range. The count is not excessive, though native_browser does add some redundancy.
The set covers the core browser interaction lifecycle: navigation, clicking, typing, scrolling, key presses, reading state, and screenshots. native_browser extends coverage to tab lifecycle, diagnostics, emulation, files, and batch actions, so the domain appears well covered.