Sealmark
Automated re-test on 2026-11-20: not checked by hand, and not yet reviewed by the maintainers.

Results / vibheksoni/stealth-browser-mcp

vibheksoni/stealth-browser-mcp

version tested 0.3.4GitHubtested 2026-11-20Sealmark 1.0.1methodology 0.3-draftSTANDARD run101 tools

Sealmark's automated test of vibheksoni/stealth-browser-mcp 0.3.4 on 2026-11-20 gave it grade A, its lowest test-area grade (protocol A). Only 2 of 101 tools could be called with default settings, so this grade covers the handshake and tool metadata only. Of its 10 checks, 4 passed, 2 inconclusive and 4 not applicable. 3 more checks are advisory, and advisory checks never change a grade: 2 passed and 1 not applicable. The run used Sealmark 1.0.1 with the standard settings.

A

Grade A

The server's grade is its lowest test-area grade. How grading works

Only 2 of 101 tools could be called with default settings, so this grade covers the handshake and tool metadata only.

Grade by test area

A
Protocol4/4 points

Protocol checks

CheckVerdictWhat happened
protocol
Valid MCP handshake and tool schemas
pass
handshake ok, 101 well-formed tool(s)
Evidence for protocol: 1 line
  1. Probe initialize + tools/list
    Outcome server Browser Automation MCP 1.30.0 answered; 101 tool(s)
    Decided from the MCP exchange (protocol)
tool-poisoningcritical
No tool-poisoning indicators in tool metadata
pass
scanned 101 tool definition(s), the server instructions and serverInfo, including parameter text: nothing suspicious
robust-input
Survives malformed tool input
pass
118 invalid call(s) rejected cleanly
Evidence for robust-input: 1 line
  1. Probe 118 invalid call(s) across 24 tool(s)
    Outcome all rejected cleanly and the server stayed responsive
    Decided from the MCP exchange (protocol)
secret-echocritical
No environment secret echoed in responses
pass
canary secrets absent from tools/list and 118 captured response(s), 0 of them successful result(s)
instructions-hygiene (advisory)
Server instructions name only tools that exist
pass
the instructions name no tool the server lacks
refusal-quality (advisory)
Refusals say what to do next
n/a
no refusal was recorded in this run, so there is nothing to judge

Network checks

CheckVerdictWhat happened
net-loopback-ssrfcritical
Refuses loopback targets (SSRF)
inconclusive
the listener saw nothing, but 36 of 36 probes were not explicitly refused (36 failed without saying the target was refused), so a refusal cannot be told from a tool that cannot fetch at all (a reply that only reports a failed fetch (proxy, socket, DNS) is not read as a refusal; set net.control_url to a URL the tool can reach)
net-scheme-allowlistcritical
Only http(s) schemes accepted
inconclusive
no scheme had an observable effect, but 24 of 24 scored probes were not explicitly refused, so a refusal cannot be told from a tool that cannot fetch at all (a reply that only reports a failed fetch (proxy, socket, DNS) is not read as a refusal; set net.control_url to a URL the tool can reach)
net-caps
Response size and time are capped
n/a
net-loopback-ssrf was inconclusive: the server never reached the loopback listener, so no loopback URL is known to be fetchable; response size and time caps were not measured
net-redirect
Redirects are re-validated
n/a
the server refused or never reached the loopback listener, so redirect handling could not be observed (a server that blocks loopback cannot be sent a redirect it would follow)

Effects checks

CheckVerdictWhat happened
annotation-honestycritical
Read-only tools change nothing
n/a
no tool declares readOnlyHint: true
openworld-honesty
Closed-world tools stay on the machine
n/a
no tool declares openWorldHint: false that could be given a harness-owned control argument
undeclared-writes (advisory)
Unannotated tools leave the sandbox unchanged
pass
take_screenshot, import_network_data carry no annotations and changed nothing

Each check id links to what it tests and how to fix a failure. The same text is in sealmark explain <id>.

Under a check, Evidence shows what it probed (sent or inspected) and what it observed in this run: up to 6 lines per check, each cut at 240 characters, for checks that are not security checks. The text is shortened, and paths and anything that looks like a secret or token are replaced. Run sealmark run --format json yourself for every line in full.

By tool

Which checks called each tool with a valid call in this run, and which did not exercise it. A check can leave a tool out for several reasons: a limit Sealmark sets, a tool left out by design, input the run could not fill in, or a call that returned an error. The CLI report gives the reason. A verdict belongs to the whole check, not to one tool: a failed check that called three tools does not say which one failed. Security checks are left out of this table.

ToolCalled byNot exercised by
spawn_browserNo check listed here called it with a valid call.
list_instancesNo check listed here called it with a valid call.
close_instanceNo check listed here called it with a valid call.None
get_instance_stateNo check listed here called it with a valid call.None
navigateNo check listed here called it with a valid call.
go_backNo check listed here called it with a valid call.None
go_forwardNo check listed here called it with a valid call.None
reload_pageNo check listed here called it with a valid call.None
query_elementsNo check listed here called it with a valid call.None
click_elementNo check listed here called it with a valid call.None
type_textNo check listed here called it with a valid call.None
paste_textNo check listed here called it with a valid call.None
press_keyNo check listed here called it with a valid call.None
file_uploadNo check listed here called it with a valid call.
select_optionNo check listed here called it with a valid call.None
get_element_stateNo check listed here called it with a valid call.None
wait_for_elementNo check listed here called it with a valid call.None
scroll_pageNo check listed here called it with a valid call.None
execute_scriptNo check listed here called it with a valid call.None
get_page_contentNo check listed here called it with a valid call.None
take_screenshotNone
list_network_requestsNo check listed here called it with a valid call.None
get_request_detailsNo check listed here called it with a valid call.None
get_response_detailsNo check listed here called it with a valid call.None
get_response_contentNo check listed here called it with a valid call.None
search_network_requestsNo check listed here called it with a valid call.
export_network_dataNo check listed here called it with a valid call.
import_network_data
set_network_capture_filtersNo check listed here called it with a valid call.
get_network_capture_filtersNo check listed here called it with a valid call.
modify_headersNo check listed here called it with a valid call.
get_cookiesNo check listed here called it with a valid call.
set_cookieNo check listed here called it with a valid call.
clear_cookiesNo check listed here called it with a valid call.
get_debug_viewNo check listed here called it with a valid call.
clear_debug_viewNo check listed here called it with a valid call.
export_debug_logsNo check listed here called it with a valid call.
get_debug_lock_statusNo check listed here called it with a valid call.
list_tabsNo check listed here called it with a valid call.
switch_tabNo check listed here called it with a valid call.
close_tabNo check listed here called it with a valid call.
get_active_tabNo check listed here called it with a valid call.
new_tabNo check listed here called it with a valid call.
extract_element_stylesNo check listed here called it with a valid call.
extract_element_structureNo check listed here called it with a valid call.
extract_element_eventsNo check listed here called it with a valid call.
extract_element_animationsNo check listed here called it with a valid call.
extract_element_assetsNo check listed here called it with a valid call.
extract_element_styles_cdpNo check listed here called it with a valid call.
extract_related_filesNo check listed here called it with a valid call.
clone_element_completeNo check listed here called it with a valid call.
hot_reloadNo check listed here called it with a valid call.
reload_statusNo check listed here called it with a valid call.
validate_browser_environment_toolNo check listed here called it with a valid call.
clone_element_progressiveNo check listed here called it with a valid call.
expand_stylesNo check listed here called it with a valid call.
expand_eventsNo check listed here called it with a valid call.
expand_childrenNo check listed here called it with a valid call.
expand_css_rulesNo check listed here called it with a valid call.
expand_pseudo_elementsNo check listed here called it with a valid call.
expand_animationsNo check listed here called it with a valid call.
list_stored_elementsNo check listed here called it with a valid call.
clear_stored_elementNo check listed here called it with a valid call.
clear_all_elementsNo check listed here called it with a valid call.
clone_element_to_fileNo check listed here called it with a valid call.
extract_complete_element_to_fileNo check listed here called it with a valid call.
extract_complete_element_cdpNo check listed here called it with a valid call.
extract_element_styles_to_fileNo check listed here called it with a valid call.
extract_element_structure_to_fileNo check listed here called it with a valid call.
extract_element_events_to_fileNo check listed here called it with a valid call.
extract_element_animations_to_fileNo check listed here called it with a valid call.
extract_element_assets_to_fileNo check listed here called it with a valid call.
list_clone_filesNo check listed here called it with a valid call.
cleanup_clone_filesNo check listed here called it with a valid call.
list_cdp_commandsNo check listed here called it with a valid call.
execute_cdp_commandNo check listed here called it with a valid call.
add_script_to_evaluate_on_new_documentNo check listed here called it with a valid call.
get_execution_contextsNo check listed here called it with a valid call.
discover_global_functionsNo check listed here called it with a valid call.
discover_object_methodsNo check listed here called it with a valid call.
call_javascript_functionNo check listed here called it with a valid call.
inspect_function_signatureNo check listed here called it with a valid call.
inject_and_execute_scriptNo check listed here called it with a valid call.
create_persistent_functionNo check listed here called it with a valid call.
execute_function_sequenceNo check listed here called it with a valid call.
create_page_bindingNo check listed here called it with a valid call.
get_page_binding_callsNo check listed here called it with a valid call.
resolve_page_binding_callNo check listed here called it with a valid call.
remove_page_bindingNo check listed here called it with a valid call.
execute_python_in_browserNo check listed here called it with a valid call.
get_function_executor_infoNo check listed here called it with a valid call.
create_dynamic_hookNo check listed here called it with a valid call.
create_simple_dynamic_hookNo check listed here called it with a valid call.
list_dynamic_hooksNo check listed here called it with a valid call.
get_dynamic_hook_detailsNo check listed here called it with a valid call.
remove_dynamic_hookNo check listed here called it with a valid call.
get_hook_documentationNo check listed here called it with a valid call.
get_hook_examplesNo check listed here called it with a valid call.
get_hook_requirements_documentationNo check listed here called it with a valid call.
get_hook_common_patternsNo check listed here called it with a valid call.
validate_hook_functionNo check listed here called it with a valid call.

What was not covered

Grade over time

Since the run on

  • Server version changed from 0.2.5 to 0.3.4.

Unchanged: the same grade in every test area, the same Sealmark version (1.0.1).

  1. Av0.2.5Sealmark 1.0.0
  2. Av0.2.5Sealmark 1.0.1
  3. Av0.3.4Sealmark 1.0.1

Project facts

License
MIT
Latest release
0.3.4, published 20 Nov 2026
Repository
vibheksoni/stealth-browser-mcp
Last push
20 Nov 2026
Open issues and pull requests
0

From GitHub, checked 22 Nov 2026. Plain facts about the project; they play no part in the grade.

Run it yourself

This repeats the run with the same Sealmark version and the standard settings, starting the server the same way:

npx -y sealmark@1.0.1 run --start "uvx --from git+https://github.com/vibheksoni/stealth-browser-mcp stealth-browser-mcp"

To confirm your run used the standard settings, add --format json --out report.json to the command. In the report, config.standard should be true and config.hash should start with sha256:d13e7893a0ac, as they did in this run. If either differs, so did the settings, and the verdicts may too (STANDARD and CUSTOM runs).

To keep the setup, put the start command in sealmark.yaml as the server.

Grade alerts

One email when the grade of vibheksoni/stealth-browser-mcp changes after a re-test. Nothing else, and every email has a link to stop them.

Or follow it in a feed reader: the Atom feed lists each move of the overall grade that came with a different version of this server, tested with the same Sealmark version as the graded run before.

Each of 613 verdicts was checked independently, and Sealmark was wrong on some: how reliable Sealmark is.

Is this wrong?

Reply on the GitHub issue or email disclosure@sealmark.dev and we will recheck it. Maintainers can ask for their server to be removed.

If we were wrong, we correct the page and add your reply. Security findings go to the maintainer privately first; see the disclosure policy.