Replay MCP lets agents inspect time-travel browser recordings, including console output, exceptions, network activity, source code, React state, performance profiles, screenshots, and execution points.

Use it from your agent

One ClawLink account serves every agent you run. Pick the guide for yours, or connect Replay once and call the same tools from all of them.

Replay for OpenClawInstall, pair, connect, first prompt

Replay for Hermes AgentInstall, pair, connect, first prompt MCPClaude Code, Cursor, any MCP clientCLI login or the remote MCP server

Setup

Connect Replay through ClawLink's hosted setup.

Auth: Paste one API key. The credential stays on the ClawLink server; your agent holds only a ClawLink key you can revoke.
Where: the ClawLink dashboard, or ask your agent to connect it.

What your agent can do

Replay exposes 32 tools through ClawLink. The 30 most used are listed here; your agent sees all of them.

Tool	What it does	Kind
`replay_mcp_annotations`	Inspect the annotations in a recording	Read
`replay_mcp_console_messages`	Show console API messages (console.log, console.warn, console.error, etc.)	Read
`replay_mcp_describe_point`	Describes the application state at a code execution point	Read
`replay_mcp_execution_delay`	Computes execution delays for functions based on their source length and the delay before the	Read
`replay_mcp_get_point_component`	Gets the React component being rendered at a code execution point	Read
`replay_mcp_get_point_link`	Generates an app.replay.io link for users to inspect a given execution point	Read
`replay_mcp_get_stack`	Gets the call stack at a code execution point	Read
`replay_mcp_inspect_element`	Get size, layout, ancestry and render point information for a DOM element	Read
`replay_mcp_list_sources`	List the sources in the recording	Read
`replay_mcp_local_storage`	Describe all the accesses made to localStorage	Read
`replay_mcp_network_request`	Analyze network requests in the recording	Read
`replay_mcp_playwright_steps`	Show all steps taken by Playwright in the recording	Read
`replay_mcp_profile_graph`	Profiles dependency graph entry execution between two points in a recording	Read
`replay_mcp_profile_sampling`	Profiles the native C++ replay engine using stack sampling between two execution points	Read
`replay_mcp_profile_statements`	Profiles JS statement execution between two points in a recording	Read
`replay_mcp_react_component_tree`	Get the structure of the committed/mounted React component tree at a point in time	Read
`replay_mcp_react_exception`	Describe any exceptions thrown which caused React to unmount	Read
`replay_mcp_react_performance_insights`	Run all deterministic React performance checks on a recording and return prioritized findings	Read
`replay_mcp_react_renders`	Analyze React renders in a recording	Read
`replay_mcp_read_source`	Read up to 50 lines of code centered around the given line and see what statements executed	Read
`replay_mcp_recording_overview`	Provides a comprehensive overview of a recording: test results, errors, environment info	Read
`replay_mcp_redux_actions`	Analyze Redux actions in a recording	Read
`replay_mcp_screenshot`	Lists all screenshot timestamps in the recording, or fetches a screenshot at a given timestamp	Read
`replay_mcp_search_sources`	Search the sources in the recording and see which statements executed	Read
`replay_mcp_tan_stack_queries`	Analyze TanStack Query (React Query) cache events in a recording	Read
`replay_mcp_uncaught_exception`	Describe uncaught exceptions and unhandled promise rejections (runtime errors not caught by	Read
`replay_mcp_user_interactions`	Show all user interactions (clicks and key presses) in the recording	Read
`replay_mcp_web_socket_messages`	Inspect the WebSocket connections in a recording and the messages sent and received on them	Read
`replay_mcp_zustand_stores`	Analyze Zustand stores in a recording	Read
`replay_mcp_warm_recording`	Starts processing a recording if needed and reports its status	Read