$ loading_
Lets coding agents receive and manage browser UI feedback.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "earmark" yet — see the docs or source repo.
Use earmark to get feedback for the clicked element in the browser overlay, including the selector, source path, and computed styles.
Returns the target element’s selector, source path, and style details for debugging.
When a control’s meaning is unclear, use earmark’s ask function to ask me a question and wait for confirmation.
Produces a clarification question so the agent can continue safely.
Please run acknowledge, resolve, and dismiss on this UI feedback item, and organize the current handling status.
Outputs updated acknowledgement, resolution, or dismissal states for the feedback item.
Frontend developers can use the browser overlay to inspect clicked elements, selectors, source paths, and computed styles to pinpoint buggy code.
When the agent cannot infer a control or state, it can ask for human confirmation and then manage the feedback flow with acknowledge, resolve, and dismiss.
Useful for automation workflows that need to track UI feedback over time, turning clicks and reviews into clear states.
It is an MCP server that gives coding agents UI feedback from a browser overlay, including clicked elements, selectors, source paths, and computed styles. It also provides watch, ask, acknowledge, resolve, and dismiss tools to manage that feedback.
It is mainly for coding agents that interact with browser UIs, and it also suits people doing frontend debugging, UI review, or automation workflows.
The provided information does not include installation or runtime details; see the source repository.
Receive and manage browser UI feedback for coding agents.
Captures recordings and turns them into timestamped Markdown bug reports.
Improve feedback-driven development with confirmations, session tracking, and prompt management.
Enable AI coding agents to communicate, share state, and coordinate work in real time.
Give AI coding agents filesystem, Git, database, and compute tools via MCP.
Create feedback-driven AI workflows with user confirmation to reduce risky actions.