Integrations / AI agent clients / CrewAI
Browser tools for your crew.
CrewAI agents take tools from MCP servers through crewai-tools. The Browz adapter starts the server inside Browz and hands its tools to an agent, so the crew’s browser work runs in isolated profiles you can watch.
pip install 'crewai[anthropic]' 'crewai-tools[mcp]' # plus the Browz Python SDK: pip install browz-mcpSetup
Set it up.
Install Browz and open it
Download Browz on the computer that runs the crew.
Install the packages
CrewAI with a model provider, the MCP tools and the Browz Python SDK.
Terminalpip install 'crewai[anthropic]' 'crewai-tools[mcp]' # plus the Browz Python SDK: pip install browz-mcpGive an agent the tools
crewai_mcp_adapter()starts the server and stops it when the block ends. SetANTHROPIC_API_KEYfirst.crew.pyfrom crewai import Agent, Crew, Task from browz.agents import AGENT_PROMPT, crewai_mcp_adapter with crewai_mcp_adapter() as tools: agent = Agent(role="Browser operator", goal="Do web tasks in Browz browsers", backstory=AGENT_PROMPT, tools=tools, llm="anthropic/claude-haiku-5-5") task = Task(description="Using Browz: create a new profile called agent-check, start it, open https://example.com, tell me the page title, take a screenshot, then stop the session and delete the profile.", expected_output="The page title", agent=agent) print(Crew(agents=[agent], tasks=[task]).kickoff())Through the hosted server (beta)
Pass
remote=True; the token comes fromBROWZ_AGENT_TOKEN.Pythonfrom browz.agents import crewai_mcp_adapter with crewai_mcp_adapter(remote=True) as tools: ... # Agent(..., tools=tools) as above
Adapter API from CrewAI MCP docs.
Example
Give the agent only some tools.
Name the tools to keep: a research agent that only reads pages does not need to type or run scripts.
from crewai import Agent, Crew, Task
from browz.agents import AGENT_PROMPT, crewai_mcp_adapter
with crewai_mcp_adapter("browz_session_open", "browz_navigate", "browz_snapshot",
"browz_screenshot", "browz_session_close") as tools:
reader = Agent(role="Page reader", goal="Read public pages and report what they say",
backstory=AGENT_PROMPT, tools=tools, llm="anthropic/claude-haiku-5-5")
task = Task(description="Open https://example.com in a throwaway Browz session and summarise it.",
expected_output="A two-sentence summary", agent=reader)
print(Crew(agents=[reader], tasks=[task]).kickoff())Good to know
Notes and tips.
- CrewAI support is in the Browz Python SDK.
- The remote server at
https://mcp.gobrowz.com/mcpis in free public beta: it needs the Browz app open with Remote control on, on a computer with a hardware security chip, and offers the profile, tab and page tools. See what it needs and what it offers. - Browz has to be installed and open on the same computer: the local MCP server drives the app’s own profiles through its local API, and reads the app’s credentials itself, so no token goes in the config.
- A person completes sign-ins, 2FA codes and CAPTCHAs. The agent can pause with
browz_request_humanand hand you the profile’s window, then carry on in the same session. Browz never solves or bypasses CAPTCHAs. - Use agents only where each site’s terms allow automated use, on accounts you are entitled to use. Our acceptable use policy applies to agents the same as to people.
- The MCP server is on every plan, Free included (5 profiles). How many agent sessions can be open at once depends on the plan: see pricing.
FAQ
CrewAI and Browz.
Do I need crewai-tools?
Yes, with its mcp extra: it provides the MCP adapter the Browz helper wraps.
Can two agents in a crew share a profile?
They can if one opens a persistent session and the other resumes it, but a session is open in one place at a time. For parallel work, give each agent its own session.
Which tool names can I pass?
Any tool of the Browz MCP server, such as browz_session_open, browz_navigate, browz_snapshot, browz_act and browz_screenshot. The full list is in the tool list.
Get started
Download Browz. Every task, its own browser.
curl -fsSL https://gobrowz.com/agents/install.sh | sh