Skip to main content
Glama
raultov

chrome-debug-mcp

by raultov
OverviewSchemaSourceRelated ServersDiscussions(1)Admin
Rust
Local

restart_chrome

Stop and restart a selected Chrome instance with remote debugging enabled to reset browser state, apply proxy settings, enable capability presets, or recover from crashes.

Instructions

Stops and restarts the Chrome instance selected by 'instance_id' (the default instance when omitted) with remote debugging enabled, optionally configuring proxy and opt-in capability presets. Side effects: destructive - terminates that instance's Chrome process and all its open tabs; closes its debugging connection; other instances keep running. Prerequisites: requires CHROME_PATH environment variable or chrome in PATH. Returns: restart success confirmation listing the presets applied. Use this to reset browser state, apply proxy settings, enable experimental browser capabilities, recover from crashes. Alternatives: 'reload' to refresh page without restart, 'navigate' to load new content. Parameters: 'features' accepts a closed set - 'WEBGL_SOFTWARE' (forces SwiftShader software WebGL for GPU-less environments); arbitrary Chrome flags are not accepted.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
featuresNoChrome capability presets to enable on the new instance. Constraints: closed set - 'WEBGL_SOFTWARE' forces SwiftShader software WebGL for GPU-less environments. Arbitrary Chrome flags are not accepted. Interactions: presets apply only to the instance started by this call and are cleared by a later restart that omits them. Defaults to: [] (no presets).
instance_idNoChrome instance id from open_instance/list_instances. Omit for the default instance.
proxy_serverNoProxy server URL (e.g., 'http://proxy.example.com:8080'). Constraints: valid proxy URL with protocol and port. Interactions: applied to new Chrome instance; requires 'enable_proxy_auth' for authenticated proxies. Defaults to: None (no proxy).

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed2 schema fields changedv1.5.1
    • changedInput schema / properties / features / description
      Previous value: -"Chrome capability presets to enable on the new instance. Constraints: closed set - 'WEB_MCP' turns on the experimental WebMCP surface for sites that expose tools to the browser; 'WEBGL_SOFTWARE' forces SwiftShader software WebGL for GPU-less environments. Arbitrary Chrome flags are not accepted. Interactions: presets apply only to the instance started by this call and are cleared by a later restart that omits them. Defaults to: [] (no presets)."New value: +"Chrome capability presets to enable on the new instance. Constraints: closed set - 'WEBGL_SOFTWARE' forces SwiftShader software WebGL for GPU-less environments. Arbitrary Chrome flags are not accepted. Interactions: presets apply only to the instance started by this call and are cleared by a later restart that omits them. Defaults to: [] (no presets)."
    • changedInput schema / properties / features / items / enum
      Previous value: -[
-  "WEB_MCP",
-  "WEBGL_SOFTWARE"
-]New value: +[
+  "WEBGL_SOFTWARE"
+]
  2. Changed6 schema fields changedv1.3.2
    • addedInput schema / properties / features / items / enum
      Added value: +[
+  "WEB_MCP",
+  "WEBGL_SOFTWARE"
+]
    • removedInput schema / properties / features / items / oneOf
      Removed value: -[
-  {
-    "enum": [
-      "WEB_MCP"
-    ]
-  },
-  {
-    "enum": [
-      "WEBGL_SOFTWARE"
-    ]
-  }
-]
    • addedInput schema / properties / features / items / type
      Added value: +"string"
    • removedInput schema / properties / features / nullable
      Removed value: -true
    • removedInput schema / properties / instance_id / nullable
      Removed value: -true
    • removedInput schema / properties / proxy_server / nullable
      Removed value: -true
  3. Changed2 schema fields changedv1.0.11
    • addedInput schema / properties / features
      Added value: +{
+  "description": "Chrome capability presets to enable on the new instance. Constraints: closed set - 'WEB_MCP' turns on the experimental WebMCP surface for sites that expose tools to the browser; 'WEBGL_SOFTWARE' forces SwiftShader software WebGL for GPU-less environments. Arbitrary Chrome flags are not accepted. Interactions: presets apply only to the instance started by this call and are cleared by a later restart that omits them. Defaults to: [] (no presets).",
+  "items": {
+    "oneOf": [
+      {
+        "enum": [
+          "WEB_MCP"
+        ]
+      },
+      {
+        "enum": [
+          "WEBGL_SOFTWARE"
+        ]
+      }
+    ]
+  },
+  "nullable": true,
+  "type": "array"
+}
    • addedInput schema / properties / instance_id
      Added value: +{
+  "description": "Chrome instance id from open_instance/list_instances. Omit for the default instance.",
+  "nullable": true,
+  "type": "string"
+}
  4. Changed1 schema field changedv1.0.7
    • addedInput schema / properties / proxy_server / description
      Added value: +"Proxy server URL (e.g., 'http://proxy.example.com:8080'). Constraints: valid proxy URL with protocol and port. Interactions: applied to new Chrome instance; requires 'enable_proxy_auth' for authenticated proxies. Defaults to: None (no proxy)."
  5. First observedv1.0.0

TDQS

A4.6/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden and does so: it declares destructive side effects (terminates the Chrome process and all open tabs, closes the debugging connection, other instances unaffected) plus a prerequisite (CHROME_PATH or chrome in PATH) and the return shape. This is exactly the behavioral context an agent needs before calling a destructive tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with the core action, then labeled sections (Side effects, Prerequisites, Returns, Alternatives, Parameters) that make it scannable. It is long, but almost every sentence carries distinct information; only the parameter restatement overlaps with the schema.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a destructive tool with no annotations and no output schema, the description covers side effects, prerequisites, return value, usage triggers, and alternatives. An agent has everything needed to call it correctly and safely.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% and the schema itself documents all three parameters in depth, including the closed-set constraint on 'features' and the proxy URL format. The description largely restates this (closed feature set, default instance) rather than adding new syntax or interaction detail, so baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('Stops and restarts the Chrome instance'), names the targeting parameter and the default-instance fallback, and distinguishes itself from siblings by naming 'reload' and 'navigate' as alternatives. An agent can tell exactly what this does without opening the schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicitly enumerates when to use it ('reset browser state, apply proxy settings, enable experimental browser capabilities, recover from crashes') and when to prefer alternatives ('reload' to refresh, 'navigate' to load new content). Nothing is left to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server

Other Tools