Featured starting point
Chrome DevTools MCP explained: browser evidence and actions for coding agents
Understand the MCP server and experimental CLI, their live-browser capabilities and the limits of a successful tool response.
Browse all 9 articles in this open-source series. Nine chapters on browser tools, setup, persistent sessions, false-value forwarding, performance evidence and responsible automation.
Featured starting point
Understand the MCP server and experimental CLI, their live-browser capabilities and the limits of a successful tool response.
Suggested learning path
01 → 09
Understand the MCP server and experimental CLI, their live-browser capabilities and the limits of a successful tool response.
Use a synthetic page and fresh target identifiers to learn a bounded browser workflow without treating successful clicks as completed tasks.
Separate server setup from browser launch, choose explicit restrictions and do not assume the experimental CLI inherits MCP defaults.
Trace persistent CLI sessions and distinguish selecting a page from isolating its browser state or restricting filesystem access.
Trace default-aware serialization and the start-path fix, including array identity and why an argv array is not a shell-escaped command.
Design controlled browser trials and distinguish telemetry latency, trace evidence, CrUX data and screenshot transport costs.
Limit browser and filesystem exposure without treating temporary profiles, header redaction or page routing as complete security boundaries.
Match integration style and debugging depth to the task without treating fewer tools as a complete permissions model.
Track targets, observations, hypotheses and verified outcomes in a read-only prototype, keeping sensitive evidence and automatic actions out of scope.