// SCAN REPORT

Looking Glass

71 tools. 49 can modify or destroy data without limits.

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

Last updated: 31 Aug 2026

49 can modify or destroy data

22 read-only

71 tools total

Community server · catalogue entry checked 05/07/2026 · full schemas captured for 46 of 71 tools

How to control Looking Glass ↓

TOOL MAP

What Looking Glass exposes to your agents

Read (22) Write / Execute (31) Destructive / Financial (8)

Running more than one server? Check your whole stack →

CONTEXT-WINDOW COST

What Looking Glass costs in tokens

5,080 tokens of tool definitions, loaded on every request

2.5% of a 200k context window

520 heaviest tool: test_scenario_run

Full per-tool breakdown → Model it in the token calculator →

MOST DANGEROUS TOOLS

Critical Risk

The most dangerous Looking Glass tools

Destructive · High browser_clear_history Clearing browser history permanently removes stored data without recovery options. Destructive · Medium browser_clear_mocks The name 'browser_clear_mocks' suggests clearing/removing mock configurations or interceptors from the browser. Destructive · High browser_clear_storage The 'clear_storage' operation implies permanent removal of data that cannot be easily recovered. Destructive · High vault_delete The name explicitly contains 'delete', which maps to the Destructive category per the classification rules.

49 of Looking Glass's 71 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control Looking Glass

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

Deny destructive operations

{
  "browser_clean_slate": {
    "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

{
  "browser_set_cookie": {
    "limits": [
      {
        "counter": "browser_set_cookie_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "browser_action_history": {
    "limits": [
      {
        "counter": "browser_action_history_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

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

Create a free account and register Looking Glass — nothing to install.
Add these rules — paste them, or build them visually. Tune the limits to your setup.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

ENFORCE POLICY ON LOOKING GLASS →

Instant setup, no code required.

FULL CATALOGUE

All 71 Looking Glass tools

DESTRUCTIVE 8 tools

Destructive browser_clean_slate browser_clean_slate Destructive browser_clear_history browser_clear_history Destructive browser_clear_mocks browser_clear_mocks Destructive browser_clear_storage browser_clear_storage Destructive browser_close browser_close Destructive session_end session_end Destructive test_chaos_clear test_chaos_clear Destructive vault_delete vault_delete

EXECUTE 29 tools

Execute browser_click browser_click Execute browser_drag browser_drag Execute browser_evaluate browser_evaluate Execute browser_explore browser_explore Execute browser_fill_and_submit browser_fill_and_submit Execute browser_go browser_go Execute browser_go_back browser_go_back Execute browser_go_forward browser_go_forward Execute browser_hover browser_hover Execute browser_login browser_login Execute browser_mock_route browser_mock_route Execute browser_navigate browser_navigate Execute browser_performance_audit browser_performance_audit Execute browser_press_key browser_press_key Execute browser_recover browser_recover Execute browser_resolve_element browser_resolve_element Execute browser_scroll browser_scroll Execute browser_select_option browser_select_option Execute browser_smart_click browser_smart_click Execute browser_tab_new browser_tab_new Execute browser_tab_select browser_tab_select Execute browser_type browser_type Execute browser_wait_for browser_wait_for Execute browser_wait_until_stable browser_wait_until_stable Execute browser_workflow browser_workflow Execute session_start session_start Execute test_chaos test_chaos Execute test_fill_form test_fill_form Execute test_scenario_run test_scenario_run

WRITE 2 tools

Write browser_set_cookie browser_set_cookie Write browser_set_localstorage browser_set_localstorage

READ 22 tools

Read browser_action_history browser_action_history Read browser_analyze_page browser_analyze_page Read browser_console_messages browser_console_messages Read browser_diff_state browser_diff_state Read browser_error_report browser_error_report Read browser_extract browser_extract Read browser_get_cookies browser_get_cookies Read browser_get_localstorage browser_get_localstorage Read browser_health_check browser_health_check Read browser_list_mocks browser_list_mocks Read browser_network_requests browser_network_requests Read browser_screenshot browser_screenshot Read browser_snapshot browser_snapshot Read browser_snapshot_state browser_snapshot_state Read browser_tab_list browser_tab_list Read session_list session_list Read session_replay session_replay Read test_accessibility_audit test_accessibility_audit Read test_scenario_status test_scenario_status Read vault_list vault_list Read visual_baseline visual_baseline Read visual_compare visual_compare

OTHER 10 tools

Other browser_diagnose browser_diagnose Other browser_suggest_actions browser_suggest_actions Other session_export_playwright session_export_playwright Other test_assert test_assert Other test_auth_flow test_auth_flow Other test_generate_assertions test_generate_assertions Other test_stop_watch test_stop_watch Other test_watch_events test_watch_events Other vault_inject vault_inject Other vault_login vault_login

EXPLORE

Related servers

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

FAQ

Questions about Looking Glass

Can an AI agent delete data through the Looking Glass MCP server? +

Yes. The Looking Glass server exposes 8 destructive tools including browser_clean_slate, browser_clear_history, browser_clear_mocks. 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 Looking Glass? +

The Looking Glass server has 2 write tools including browser_set_cookie, browser_set_localstorage. 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 Looking Glass.

How many tools does the Looking Glass MCP server expose? +

71 tools across 4 categories: Destructive, Execute, Read, Write. 22 are read-only. 49 can modify, create, or delete data.

How do I enforce a policy on Looking Glass? +

Register the Looking Glass 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 Looking Glass tool call.

Deterministic rules across all 71 Looking Glass tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON LOOKING GLASS →

Instant setup, no code required.

71 Looking Glass tools catalogued and risk-classified — across an index of 46,500+ MCP servers.

// WHERE THIS COMES FROM

These policies come from Looking Glass's registry record.

The record behind this page: verified identity, auth posture, risk grade, every tool classified, recommended policy — re-checked continuously.

Teams ship this data inside their own products. See what a licence covers →