UI Bridge MCP

149 tools. 68 can modify or destroy data without limits.

5 destructive tools with no built-in limits. Policy required.

Last updated:

68 can modify or destroy data
81 read-only
149 tools total

Community server · catalogue entry verified 02/07/2026

How to control UI Bridge MCP ↓

What UI Bridge MCP exposes to your agents

Read (81) Write / Execute (63) Destructive / Financial (5)
Critical Risk

The most dangerous UI Bridge MCP tools

68 of UI Bridge MCP's 149 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

How to control UI Bridge MCP

PolicyLayer is an MCP gateway — it sits between your AI agents and UI Bridge MCP, and nothing reaches the server without passing your rules. These are the rules we recommend:

Deny destructive operations
{
  "sdk_clear": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations
{
  "sdk_clipboard_write": {
    "limits": [
      {
        "counter": "sdk_clipboard_write_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations
{
  "diagnose_stuck_screen": {
    "limits": [
      {
        "counter": "diagnose_stuck_screen_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

  1. Create a free account and register UI Bridge MCP — nothing to install.
  2. Add these rules — paste them, or build them visually. Tune the limits to your setup.
  3. Point your MCP client (Claude, Cursor, anything) at your gateway URL.
ENFORCE POLICY ON UI BRIDGE →

Instant setup, no code required.

All 149 UI Bridge MCP tools

EXECUTE 50 tools
Execute sdk_ai_execute sdk_ai_execute Execute sdk_blur Remove focus from an element in the SDK app. Execute sdk_change_buffer_disable Disable change buffering. Execute sdk_change_buffer_enable Enable change buffering. All diffs are accumulated until drained. Execute sdk_click sdk_click Execute sdk_double_click Double-click an element in the SDK app. Execute sdk_drag Drag an element to a target in the SDK app. Execute sdk_execute_with_diff Execute an element action and capture what changed in the UI. Returns before/after diff with categorization. Execute sdk_focus Focus an element in the SDK app. Execute sdk_hover Hover over an element in the SDK app. Execute sdk_media_audit_performance sdk_media_audit_performance Execute sdk_page_go_back sdk_page_go_back Execute sdk_page_go_forward sdk_page_go_forward Execute sdk_page_navigate sdk_page_navigate Execute sdk_page_refresh sdk_page_refresh Execute sdk_redo Execute redo in the connected SDK app. Uses the app Execute sdk_right_click Right-click an element in the SDK app. Execute sdk_scroll Scroll within an element in the SDK app. Execute sdk_select Select an option in a dropdown in the SDK app. Execute sdk_submit Submit the form containing the element in the SDK app. Execute sdk_toggle Toggle a checkbox in the SDK app. Execute sdk_type sdk_type Execute sdk_undo Execute undo in the connected SDK app. Uses the app Execute sdk_wait_for_change Wait for the UI to change matching a predicate. Polls until a matching change is detected or timeout. Execute sdk_wait_for_network_request sdk_wait_for_network_request Execute ui_blur Remove focus from an element in the runner Execute ui_change_buffer_drain Drain the change buffer in the runner Execute ui_check Check a checkbox element in the runner Execute ui_click ui_click Execute ui_double_click Double-click an element in the runner Execute ui_drag ui_drag Execute ui_execute_with_diff Execute an element action in the runner Execute ui_fill_form ui_fill_form Execute ui_focus Focus an element in the runner Execute ui_hover Hover over an element in the runner Execute ui_page_evaluate ui_page_evaluate Execute ui_redo Execute redo in the runner Execute ui_right_click Right-click an element in the runner Execute ui_scroll Scroll within an element in the runner Execute ui_select Select an option in a dropdown/select element in the runner Execute ui_submit Submit the form containing the element in the runner Execute ui_toggle Toggle a checkbox element in the runner Execute ui_type ui_type Execute ui_uncheck Uncheck a checkbox element in the runner Execute ui_undo Execute undo in the runner Execute ui_wait_for_change Wait for the runner Execute ui_wait_for_network_request ui_wait_for_network_request Execute wait_for_idle wait_for_idle Execute wait_for_signal wait_for_signal Execute wait_for_targets wait_for_targets
READ 81 tools
Read diagnose_stuck_screen diagnose_stuck_screen Read get_idle_status get_idle_status Read sdk_ai_assert sdk_ai_assert Read sdk_ai_search sdk_ai_search Read sdk_analyze_data sdk_analyze_data Read sdk_analyze_regions sdk_analyze_regions Read sdk_analyze_structured_data sdk_analyze_structured_data Read sdk_annotated_screenshot sdk_annotated_screenshot Read sdk_categorize_last_diff Categorize the last computed diff. Returns category Read sdk_change_buffer_drain Drain the change buffer, returning all accumulated changes since last drain. Read sdk_change_buffer_size Get the current change buffer size and enabled status. Read sdk_check Check a checkbox in the SDK app. Read sdk_clipboard_read Read the current system clipboard text content (SDK mode). Read sdk_cross_app_compare sdk_cross_app_compare Read sdk_design_audit sdk_design_audit Read sdk_design_diff sdk_design_diff Read sdk_design_evaluate sdk_design_evaluate Read sdk_design_load_guide sdk_design_load_guide Read sdk_design_responsive sdk_design_responsive Read sdk_design_review sdk_design_review Read sdk_design_state_styles sdk_design_state_styles Read sdk_design_styles sdk_design_styles Read sdk_diff sdk_diff Read sdk_diff_from_bookmark Compare current UI state against a saved bookmark. Returns appeared, disappeared, and modified elements. Read sdk_discover sdk_discover Read sdk_element_screenshot sdk_element_screenshot Read sdk_elements sdk_elements Read sdk_find sdk_find Read sdk_form_diff sdk_form_diff Read sdk_form_snapshot sdk_form_snapshot Read sdk_forms sdk_forms Read sdk_get_bookmark Get a specific bookmark Read sdk_get_element sdk_get_element Read sdk_list_bookmarks List all saved snapshot bookmarks. Read sdk_media_analyze_batch sdk_media_analyze_batch Read sdk_media_audit_accessibility sdk_media_audit_accessibility Read sdk_media_compare sdk_media_compare Read sdk_media_find sdk_media_find Read sdk_media_snapshot sdk_media_snapshot Read sdk_network_requests sdk_network_requests Read sdk_network_requests_in_flight sdk_network_requests_in_flight Read sdk_page_summary sdk_page_summary Read sdk_scoped_diff Get a diff scoped to a CSS selector region. Optionally diff from a bookmark. Read sdk_screenshot sdk_screenshot Read sdk_screenshot_after sdk_screenshot_after Read sdk_screenshot_baseline_list sdk_screenshot_baseline_list Read sdk_screenshot_before sdk_screenshot_before Read sdk_screenshot_delta sdk_screenshot_delta Read sdk_screenshot_diff sdk_screenshot_diff Read sdk_snapshot sdk_snapshot Read sdk_status sdk_status Read sdk_structured_changes Analyze table and list changes between snapshots. Detects added/removed rows and list items. Read sdk_summarize_diff Get a budget-aware text summary of UI changes. Summarizes appeared/disappeared/modified elements within a char Read sdk_uncheck Uncheck a checkbox in the SDK app. Read sdk_undo_state Get undo/redo availability and state from the connected SDK app. Shows whether undo/redo is available, what it Read sdk_visual_description sdk_visual_description Read ui_analyze_image ui_analyze_image Read ui_annotated_screenshot ui_annotated_screenshot Read ui_audit_images ui_audit_images Read ui_categorize_last_diff Categorize the last computed diff in the runner Read ui_change_buffer_size Get the change buffer size and enabled status in the runner Read ui_clipboard_read Read the current system clipboard text content (Control mode). Read ui_diff ui_diff Read ui_diff_from_bookmark Compare current runner UI state against a saved bookmark (Control mode). Returns appeared, disappeared, and mo Read ui_discover ui_discover Read ui_element_screenshot ui_element_screenshot Read ui_form_diff ui_form_diff Read ui_form_snapshot ui_form_snapshot Read ui_forms ui_forms Read ui_get_bookmark Get a specific bookmark Read ui_get_element ui_get_element Read ui_health Check if the qontinui-runner is running and accessible. Read ui_list_bookmarks List all saved snapshot bookmarks in the runner Read ui_network_requests ui_network_requests Read ui_network_requests_in_flight ui_network_requests_in_flight Read ui_query_selector ui_query_selector Read ui_scoped_diff Get a diff scoped to a CSS selector region in the runner Read ui_snapshot ui_snapshot Read ui_structured_changes Analyze table and list changes in the runner Read ui_summarize_diff Get a budget-aware text summary of runner UI changes (Control mode). Read ui_undo_state Get undo/redo availability and state in the runner

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

Questions about UI Bridge MCP

Can an AI agent delete data through the UI Bridge MCP server? +

Yes. The UI Bridge MCP server exposes 5 destructive tools including sdk_clear, sdk_delete_bookmark, sdk_reset. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through UI Bridge MCP? +

The UI Bridge MCP server has 13 write tools including sdk_clipboard_write, sdk_connect, sdk_disconnect. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach UI Bridge MCP.

How many tools does the UI Bridge MCP server expose? +

149 tools across 4 categories: Destructive, Execute, Read, Write. 81 are read-only. 68 can modify, create, or delete data.

How do I enforce a policy on UI Bridge MCP? +

Register the UI Bridge MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Enforce policy on every UI Bridge MCP tool call.

Deterministic rules across all 149 UI Bridge MCP tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

Instant setup, no code required.

149 UI Bridge MCP tools catalogued and risk-classified — across an index of 43,000+ MCP servers.

// GET IN TOUCH

Have a question or want to learn more? Send us a message.

Message sent.

We'll get back to you soon.