This MCP Browser Agent server enables automated browser interactions through Claude using the Model Context Protocol. It provides browser automation capabilities for navigation and complex workflows, dynamic state management with planning context that includes evaluation of previous goals and memory of past actions, and action execution tools to implement sequences of planned browser actions. The server offers interactive element detection and manipulation on web pages, session recording in the ./tmp/recordings directory, and AI integration for intelligent browser control and decision-making. It supports configurable browser contexts with customizable settings and includes comprehensive logging and debugging features for development.
Provides automated browser interaction within Google Chrome, allowing for element detection, navigation, and state management in web applications
Supports browser automation on macOS systems with specific compatibility for darwin 24.2.0
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@MCP Browser Agent search for flights from New York to London next week"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Built at AGI House MCP Hackathon
This project is a browser automation agent that uses the Model Context Protocol (MCP) to enable browser interactions. It provides a seamless integration between Claude and browser automation capabilities through our MCP server.
Thank you to Browser-Use for their browser agent capabilities that help power our MCP server!
Related MCP server: selenium-mcp
macOS (darwin 24.2.0)
Python 3.12 or higher
uv package manager
Google Chrome browser (Ensure your browser is closed before running task(s).)
To install Browser Automation Agent for Claude Desktop automatically via Smithery:
npx -y @smithery/cli install @ashley-ha/mcp-manus --client claudeClone the repository:
git clone <repository-url>
cd mcpSet up the Python environment using uv:
uv venv
source .venv/bin/activate
uv syncCreate or modify your Claude Desktop configuration file:
{
"mcpServers": {
"browser-use": {
"command": "uv",
"args": [
"--directory",
"/ABSOLUTE/PATH/TO/mcp",
"run",
"browser-use.py"
]
}
}
}Replace /ABSOLUTE/PATH/TO/browser-use with the absolute path to your project directory.
The agent is configured to use Google Chrome with the following default settings:
Non-headless mode for development
Window size: 1280x1100
Disabled security features for testing
Recording path: ./tmp/recordings
Browser automation through MCP tools
State management and planning capabilities
Interactive element detection and manipulation
Configurable browser contexts
Logging and debugging support
The agent provides two main tools:
get_planner_state: Retrieves the current browser state and planning context
execute_actions: Executes planned actions in the browser
The project uses Python's built-in logging with the following configuration:
All logs are directed to stderr
Custom formatting: %(levelname)-8s [%(name)s] %(message)s
Root logger level: INFO
Third-party loggers level: WARNING
browser-use.py: Main entry point and server implementation
tmp/recordings: Directory for browser session recordings
Dependencies managed through uv
This project was built during the AGI House MCP Hackathon. Contributions are welcome!
This project is licensed under the MIT License - see the LICENSE file for details.
Copyright (c) 2025 Jaeyun Ha, Ashley Ha
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
Execute actions from the planner state.
Args:
actions: A dictionary containing the planner state and actions in format:
{
"current_state": {
"evaluation_previous_goal": str,
"memory": str,
"next_goal": str
},
"action": [
{"action_name": {"param1": "value1"}},
...
]
}
Note: If the page state changes (new elements appear) during action execution,
the sequence will be interrupted and you'll need to get a new planner state.
| Name | Required | Description | Default |
|---|---|---|---|
| actions | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden. It discloses that execution can be interrupted by page state changes, which is a key behavioral trait, but doesn't cover other aspects like error handling, side effects, or response format. It adds some context but is incomplete for a mutation tool.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is appropriately sized and front-loaded with the main purpose, followed by an 'Args' section and a note. The structure is clear, but the note could be more integrated; overall, it's efficient with minimal waste.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the complexity (1 parameter with nested objects, no annotations, no output schema), the description covers the parameter structure well and includes a behavioral note. However, it lacks details on return values, error cases, and full usage context, making it adequate but with gaps for a tool that likely performs mutations.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema description coverage is 0%, so the description must compensate. It provides a detailed example of the 'actions' parameter structure, including nested objects and keys like 'current_state' and 'action', which adds significant meaning beyond the schema's generic 'object' type. This effectively documents the parameter semantics.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states the tool 'Execute actions from the planner state', which provides a verb ('Execute') and resource ('actions from the planner state'), but it's vague about what 'actions' specifically entail (e.g., UI interactions, API calls) and doesn't clearly distinguish from the sibling tool 'get_planner_state'. It's not tautological but lacks specificity.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description includes a note about interruption when 'the page state changes', which implies a usage context (e.g., web automation), but it doesn't explicitly state when to use this tool versus alternatives like 'get_planner_state' or provide prerequisites. The guidance is minimal and not comprehensive.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get the current browser state and planning context. This tool must be executed before execute_actions tool.
Must return a JSON string in the format:
{
"current_state": {
"evaluation_previous_goal": "Success|Failed|Unknown - Analysis of previous actions",
"memory": "Description of what has been done and what to remember",
"next_goal": "What needs to be done with the next immediate action"
},
"action": [
{"action_name": {"param1": "value1", ...}},
...
]
}
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden of behavioral disclosure. It effectively describes key behavioral traits: it's a read operation ('Get'), it returns specific structured data (a JSON string with defined format), and it has a prerequisite relationship with another tool. It doesn't cover aspects like error handling or performance, but for a zero-parameter tool with no annotations, this is reasonably comprehensive.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is front-loaded with the core purpose and usage guideline in the first two sentences, which is good. However, it includes a detailed JSON format specification that might be better suited for an output schema. While this adds value, it makes the description longer than necessary for conciseness, as the output details could be separated into structured data.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool has 0 parameters, no annotations, and no output schema, the description provides good contextual completeness. It explains the purpose, usage guidelines, and output format in detail. The only gap is the lack of an output schema, but the description compensates by specifying the return format explicitly, making it sufficient for the agent to understand how to use the tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool has 0 parameters with 100% schema description coverage, so the baseline is 4. The description doesn't need to add parameter information, and it doesn't attempt to, which is appropriate. No parameters are present to document, so this score reflects that the description doesn't introduce confusion or redundancy regarding inputs.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose: 'Get the current browser state and planning context.' It specifies the verb ('Get') and resource ('browser state and planning context'), making it easy to understand what the tool does. However, it doesn't explicitly differentiate from its sibling tool 'execute_actions' beyond stating a prerequisite relationship, which is more about usage than purpose distinction.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides explicit usage guidance: 'This tool must be executed before execute_actions tool.' It clearly states when to use this tool (as a prerequisite for 'execute_actions') and implies an alternative (use 'execute_actions' after this). This is a strong, directive guideline that helps the agent understand the tool's role in the workflow.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
execute_actionsget_planner_stateScored across 2 tools
The two tools have completely distinct purposes: get_planner_state retrieves browser state and planning context, while execute_actions performs actions based on that state. There is no overlap or ambiguity between these functions.
Both tools follow a consistent verb_noun pattern with clear action-oriented names (get_planner_state, execute_actions). The naming convention is uniform and predictable throughout the set.
With only 2 tools for a browser automation server, the surface feels severely limited. While the tools cover a basic planning-execution loop, typical browser automation requires more granular operations like navigation, element interaction, or content extraction.
The toolset provides only a high-level planning/execution abstraction without direct browser manipulation capabilities. There are significant gaps for common browser tasks like navigating to URLs, clicking elements, extracting text, or handling dialogs, which agents would need for robust automation.
AI-powered web automation. Navigate websites using AI agents for one page or a thousand
AI-powered web automation. Navigate websites using AI agents for one page or a thousand
Autonomous cloud agent tasks: real browser + your tools, structured evidence-backed results.
AI-powered browser automation — navigate, click, fill forms, and extract data from any website.
