MCP Browser
NPM · MCPBROWSER · SCANNED SEP 22
Browser automation using the user's Chromium session — JS execution, auth, and anti-bot support.
Available components
How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →
Supply Chain Security98
- No malware found by supply-chain analysis.Pass
- A known CVE affects extract-zip, but no fixed version has been published, so there is nothing to upgrade to. View diagnostics → Partial
- No install/post-install scripts declared.Pass
- 55 of 174 dependencies flagged as unhealthy. View diagnostics → Partial
Provenance & Transparency45
- Source repository is publicly reachable at the declared URL. View diagnostics → Pass
- Provenance check failed: no build-provenance attestation is published. See how to fix → View diagnostics → Fail
- Clear OSI-approved license (MIT).Pass
- Actively maintained (last published 41 days ago).Pass
- Disclosure check failed: no security disclosure policy was found in the source repository. See how to fix → Fail
Schema Quality & AI Usability80
- 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).Pass
- AI-judged instruction clarity (excellent).Pass
- Context-footprint check failed: tool/resource definitions use about 1967 tokens (~196/item across 10 items; 10 tools + 0 resources), over budget; trim descriptions and params. See how to fix → Fail
- Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management90
- Stability observed for 27 of 30 days with no destabilising changes; credit accrues until the full window elapses.Partial
Tool Coverage100
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 100% of tool parameters carry a description.Pass
- Structured output schemas are declared (100% of tools); any adoption earns full credit.Pass
Tool Safety100
- No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
- All 1 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation.Pass
- An AI judge read all 11 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
- Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.Pass
How do I install the MCP Browser server?
MCP Browser runs locally as an npm package, launched with npx -y mcpbrowser. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
npm · mcpbrowser
claude mcp add cherchyk-mcpbrowser -- npx -y mcpbrowser
{
"mcpServers": {
"cherchyk-mcpbrowser": {
"command": "npx",
"args": [
"-y",
"mcpbrowser"
]
}
}
} {
"servers": {
"cherchyk-mcpbrowser": {
"command": "npx",
"args": [
"-y",
"mcpbrowser"
]
}
}
} codex mcp add cherchyk-mcpbrowser -- npx -y mcpbrowser
{
"$schema": "https://opencode.ai/config.json",
"mcp": {
"cherchyk-mcpbrowser": {
"type": "local",
"command": [
"npx",
"-y",
"mcpbrowser"
],
"enabled": true
}
}
} openclaw mcp add cherchyk-mcpbrowser --command npx --arg -y --arg mcpbrowser
mcp_servers:
cherchyk-mcpbrowser:
command: "npx"
args: ["-y", "mcpbrowser"] {
"McpServers": {
"cherchyk-mcpbrowser": {
"Transport": "stdio",
"Command": "npx",
"Arguments": [
"-y",
"mcpbrowser"
]
}
}
} assistant mcp add cherchyk-mcpbrowser -t stdio -c npx -a -y mcpbrowser
{
"mcpServers": {
"cherchyk-mcpbrowser": {
"command": "npx",
"args": [
"-y",
"mcpbrowser"
]
}
}
} Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.
- 22 Sept 26 0
- Security disclosure: unverified → fail ▼ functional
- 21 Sept 26 +1
- Security disclosure: fail → unverified ▼ functional
- 19 Sept 26 −3
- Stability: pass → 0.80 functional
- 18 Sept 26 +1
- Stability: 0.97 → pass security
- 16 Sept 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 90 to 93. That category is still filling its 30-day observation window: 27 days of observed history at the previous scan, 28 at this one. The score rises as the window fills, whether or not the server changes.
- 14 Sept 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 83 to 87. That category is still filling its 30-day observation window: 25 days of observed history at the previous scan, 26 at this one. The score rises as the window fills, whether or not the server changes.
- 12 Sept 26 −3
- Stability: pass → 0.80 functional
- 10 Sept 26 +1
- Stability: 0.97 → pass security
Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.
Captured 22 Sept 2026 · Analysed npm/mcpbrowser@0.4.6
Provenance No attestation
The registry publishes no build provenance for this version, so there is nothing to verify.
| Result | No attestation |
|---|---|
| Ecosystem | npm |
Background: How many MCP packages publish verified provenance →
Vulnerabilities 2 findings
| ID | CVE | Severity | Vector | Fix available |
|---|---|---|---|---|
| GHSA-7pqw-9j4j-h8q3 | CVE-2026-19693 | high | CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:H/A:H | no |
| GHSA-jmr9-qjv8-65gv | CVE-2026-56876 | high | CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:N | no |
Background: What a vulnerability scan can and cannot prove →
Dependencies 174 packages
| Packages resolved | 174 |
|---|---|
| Stale | 55 |
| Tree resolution | Complete |
Background: SBOMs and build attestations, explained →
The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →
browser_click_element Click Element ~208
Click buttons, links, or any element on a browser-loaded page. Use when: you need to navigate a website, submit a form, press a button, follow a link, or interact with any clickable UI element. Targets by CSS selector or visible text. Returns updated page HTML after click. PREREQUISITE: Page must be loaded with browser_fetch_webpage first.
| Name | Type | Req | Description |
|---|---|---|---|
| detectForms | boolean | – | Scan page for forms after click and return structured form data (fields, selectors, submit buttons, orphaned inputs). Only applies when returnHtml=true. Set to true when you need to fill or interact… |
| selector | string | – | CSS selector for the element to click (e.g., '#submit-btn', '.login-button') |
| text | string | – | Text content to search for if selector is not provided (e.g., 'Sign In', 'Submit') |
| url | string | yes | The URL of the page (must match a previously fetched page) |
| Name | Type | Req | Description |
|---|---|---|---|
| currentUrl | string | yes | URL after click |
| fallbackAttempt | object | – | Present when fallbackUsed is true |
| fallbackUsed | boolean | yes | True when native click timed out and JS fallback ran |
| forms | array | – | Detected forms with fields, selectors, and metadata (when returnHtml is true) |
| html | string | yes | Page HTML if returnHtml was true, null otherwise |
| message | string | yes | Status message |
| nativeAttempt | object | yes | – |
| nextSteps | array | yes | Suggested next actions |
| orphanedFields | array | – | Input/select/textarea elements not inside any <form> (when returnHtml is true) |
| postClickWait | object | – | Post-click wait metadata |
| recommendedPlugins | array | – | Detected site-specific plugins available for this domain |
| scrollableAreas | array | – | Scrollable containers on the page. Pass a selector to browser_scroll_page's 'container' parameter to scroll within a specific area. |
| status | string | yes | Overall click status after native and fallback attempts |
| totalFieldCount | number | – | Total number of form fields found on the page |
No examples provided.
browser_close_tab Close Tab ~62
Close a browser tab to free resources. Use when: you are done with a page and want to release memory, or need to reset session state for a hostname. Uses exact hostname match.
| Name | Type | Req | Description |
|---|---|---|---|
| url | string | yes | The URL whose hostname tab should be closed |
| Name | Type | Req | Description |
|---|---|---|---|
| hostname | string | yes | Hostname that was closed |
| message | string | yes | Success message |
| nextSteps | array | yes | Suggested next actions |
No examples provided.
browser_detect_forms Detect Forms ~95
Scan a browser-loaded page and return all forms as structured JSON — fields, types, validation rules, submit buttons, and orphaned inputs. Use when: you need to understand a form before filling it, discover what fields exist on a page, or map form structure for automation. PREREQUISITE: Page must be loaded with browser_fetch_webpage first.
| Name | Type | Req | Description |
|---|---|---|---|
| url | string | yes | URL of the already-loaded page |
| Name | Type | Req | Description |
|---|---|---|---|
| forms | array | yes | Array of detected forms with fields and metadata |
| nextSteps | array | yes | Suggested next actions |
| orphanedFields | array | yes | Input/select/textarea elements not inside any <form> |
| summary | string | yes | Human-readable summary of detected forms |
| totalFieldCount | number | yes | Total number of fields found |
No examples provided.
browser_execute_javascript Execute JavaScript ~125
Run JavaScript on a browser-loaded page and get the result. Use when: you need to extract structured data from a page, manipulate the DOM, read page state, run custom queries on page content, or perform UI actions that CSS selectors cannot reach. Returns the script result as JSON, text, or void. PREREQUISITE: Page must be loaded with browser_fetch_webpage first.
| Name | Type | Req | Description |
|---|---|---|---|
| script | string | yes | JavaScript source code to execute in page context |
| url | string | yes | The URL of the page (must match a previously fetched page) |
| Name | Type | Req | Description |
|---|---|---|---|
| currentUrl | string | yes | URL after execution |
| error | object | – | Error object when script throws or times out |
| executionTimeMs | number | yes | Script execution duration |
| nextSteps | array | yes | – |
| recommendedPlugins | array | – | Detected site-specific plugins available for this domain |
| result | string | – | Serialized result of the script |
| truncated | boolean | yes | True if result was capped to size limit |
| type | string | yes | Type of the returned result |
| urlChanged | boolean | yes | True if page URL changed during execution |
No examples provided.
browser_fetch_webpage Fetch Web Page ~305
Load any URL and return its content using a real browser with full JavaScript execution. Use whenever a webpage needs to be loaded — whether to read content, interact with a web app, or access any HTTP/HTTPS resource. Connects to the user's existing browser session with all cookies, logins, and SSO sessions available automatically. Handles JavaScript rendering, authentication, and anti-bot protection. Prefer this over generic HTTP fetch for authenticated or JS-heavy pages.
| Name | Type | Req | Description |
|---|---|---|---|
| detectForms | boolean | – | Scan page for forms and return structured form data (fields, selectors, submit buttons, orphaned inputs). Set to true when you need to fill or interact with forms. |
| format | string | – | Format for the returned content. 'html' (default) returns cleaned HTML. 'text' returns normalized visible text. 'markdown' returns readable Markdown. For 'text' and 'markdown' without a selector, MCP… |
| selector | string | – | CSS selector to extract a specific DOM subtree instead of the full page. Prefer semantic content regions like 'main', 'article', or '[role="main"]' to capture the primary content while skipping navig… |
| url | string | yes | The URL to fetch |
| Name | Type | Req | Description |
|---|---|---|---|
| currentUrl | string | yes | Final URL after any redirects |
| forms | array | – | Detected forms with fields, selectors, and metadata |
| html | string | yes | Page content in the requested format (cleaned HTML by default; visible text or Markdown when 'format' is set) |
| nextSteps | array | yes | Suggested next actions |
| orphanedFields | array | – | Input/select/textarea elements not inside any <form> |
| recommendedPlugins | array | – | Detected site-specific plugins available for this domain |
| scrollableAreas | array | – | Scrollable containers on the page. Pass a selector to browser_scroll_page's 'container' parameter to scroll within a specific area. |
| totalFieldCount | number | – | Total number of form fields found on the page |
No examples provided.
browser_get_current_html Get Current HTML ~295
Re-read HTML from an already-loaded page without reloading it. Use when: you need to check page state after a click or form fill, re-extract content from the current page, or get updated HTML after dynamic changes. Much faster than browser_fetch_webpage since it skips navigation. PREREQUISITE: Page must be loaded with browser_fetch_webpage first.
| Name | Type | Req | Description |
|---|---|---|---|
| detectForms | boolean | – | Scan page for forms and return structured form data (fields, selectors, submit buttons, orphaned inputs). Set to true when you need to fill or interact with forms. |
| format | string | – | Format for the returned content. 'html' (default) returns cleaned HTML. 'text' returns normalized visible text. 'markdown' returns readable Markdown. For 'text' and 'markdown' without a selector, MCP… |
| selector | string | – | CSS selector to extract a specific DOM subtree instead of the full page. Prefer semantic content regions like 'main', 'article', or '[role="main"]' to capture the primary content while skipping navig… |
| url | string | yes | The URL of the page (must match a previously fetched page) |
| Name | Type | Req | Description |
|---|---|---|---|
| currentUrl | string | yes | Current page URL |
| forms | array | – | Detected forms with fields, selectors, and metadata |
| html | string | yes | Page content in the requested format (cleaned HTML by default; visible text or Markdown when 'format' is set) |
| nextSteps | array | yes | Suggested next actions |
| orphanedFields | array | – | Input/select/textarea elements not inside any <form> |
| recommendedPlugins | array | – | Detected site-specific plugins available for this domain |
| scrollableAreas | array | – | Scrollable containers on the page. Pass a selector to browser_scroll_page's 'container' parameter to scroll within a specific area. |
| totalFieldCount | number | – | Total number of form fields found on the page |
No examples provided.
browser_navigate_history Navigate Back/Forward ~107
Go back or forward in browser history. Use when: you clicked a link and need to return to the previous page, or want to go forward after going back. PREREQUISITE: Page must be loaded with browser_fetch_webpage first.
| Name | Type | Req | Description |
|---|---|---|---|
| direction | string | – | Navigation direction: 'back' to go to previous page, 'forward' to go to next page |
| url | string | yes | URL of the already-loaded page (identifies which tab to navigate) |
| Name | Type | Req | Description |
|---|---|---|---|
| currentUrl | string | yes | URL after navigation |
| direction | string | yes | Navigation direction used |
| html | string | – | Page HTML content after navigation (null if returnHtml=false) |
| nextSteps | array | yes | Suggested next actions |
| previousUrl | string | yes | URL before navigation |
No examples provided.
browser_scroll_page Scroll Page ~313
Scroll within a browser-loaded page. Automatically detects scrollable containers inside SPAs (e.g., ADO, Jira, Gmail) where the main content scrolls inside an inner div rather than the window. Use when: you need to see more content below the fold, bring an element into view before clicking, scroll to a specific section, or navigate long pages. Supports scroll by direction, to a CSS selector, or to absolute coordinates. PREREQUISITE: Page must be loaded with browser_fetch_webpage first.
| Name | Type | Req | Description |
|---|---|---|---|
| amount | number | – | Pixels to scroll in the specified direction. Default: 500 (roughly half a viewport). |
| container | string | – | CSS selector of a specific scrollable container to scroll within. Use when the page has multiple scroll areas (e.g., a sidebar + main content). Get available selectors from the scrollableAreas field… |
| direction | string | – | Direction to scroll. Use with 'amount' parameter. |
| selector | string | – | CSS selector of element to scroll into view. When provided, ignores direction/amount and scrolls until the element is visible. |
| url | string | yes | The URL of the page (must match a previously fetched page) |
| x | number | – | Absolute horizontal scroll position. Use with 'y' for precise positioning. |
| y | number | – | Absolute vertical scroll position. Use with 'x' for precise positioning. |
| Name | Type | Req | Description |
|---|---|---|---|
| currentUrl | string | yes | Current page URL |
| nextSteps | array | yes | Suggested next actions |
| pageHeight | number | yes | Total scrollable page height |
| pageWidth | number | yes | Total scrollable page width |
| scrollX | number | yes | New horizontal scroll position in pixels |
| scrollY | number | yes | New vertical scroll position in pixels |
| scrollableAreas | array | – | Scrollable containers detected on the page. Pass a selector to the 'container' parameter to scroll within a specific area. |
| viewportHeight | number | yes | Visible viewport height |
| viewportWidth | number | yes | Visible viewport width |
No examples provided.
browser_take_screenshot Take Screenshot ~137
Capture a screenshot of a browser-loaded page as PNG. Set fullPage=true to capture the entire scrollable page in one shot — this avoids multiple scroll+screenshot cycles. Only use fullPage=false when you specifically need just the current viewport (rare). Use when: you need to see what a page looks like, analyze visual layout, view charts/images/graphs, debug UI issues, or when HTML alone is insufficient. Returns base64-encoded PNG. PREREQUISITE: Page must be loaded with browser_fetch_webpage first.
| Name | Type | Req | Description |
|---|---|---|---|
| url | string | yes | The URL of the page (must match a previously fetched page) |
| Name | Type | Req | Description |
|---|---|---|---|
| currentUrl | string | yes | Current page URL |
| mimeType | string | yes | Image MIME type (image/png) |
| nextSteps | array | yes | Suggested next actions |
| screenshotBase64 | string | yes | Base64-encoded PNG screenshot |
No examples provided.
browser_type_text Type Text ~116
Type text into input fields on a browser-loaded page. Use when: you need to fill a form, enter a search query, type into a text box, or input data into any editable field. Supports filling multiple fields in a single call. PREREQUISITE: Page must be loaded with browser_fetch_webpage first.
| Name | Type | Req | Description |
|---|---|---|---|
| fields | array | yes | Array of fields to fill. Each field specifies a selector and text to type. |
| url | string | yes | The URL of the page (must match a previously fetched page) |
| Name | Type | Req | Description |
|---|---|---|---|
| currentUrl | string | yes | URL after typing |
| html | string | yes | Page HTML if returnHtml was true, null otherwise |
| message | string | yes | Success message |
| nextSteps | array | yes | Suggested next actions |
No examples provided.
What is the MCP Browser server?
MCP Browser is listed in the public MCP registry as io.github.cherchyk/mcpbrowser. Browser automation using the user's Chromium session, JS execution, auth, and anti-bot support. This page covers its npm package (mcpbrowser).
Is the MCP Browser server safe to use?
MCP Browser scores 83 out of 100 on VerifyMCP. We recorded 2 known advisories against it as of 22 September 2026. It declares no install or post-install scripts. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.
What tools does the MCP Browser server expose?
MCP Browser exposes 10 tools: browser_fetch_webpage, browser_execute_javascript, browser_click_element, browser_type_text, browser_close_tab, and 5 more. Their descriptions and schemas cost roughly 1,763 tokens of context every time the server is loaded.
Is the MCP Browser server still maintained?
MCP Browser is still listed as active in the MCP registry. We last reached this channel on 22 September 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.
What licence is the MCP Browser server under?
MCP Browser declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.