Extracts academic PDF content from arXiv papers, supporting abstract mode (~1500 chars) and full text extraction.
Provides Google search capabilities, including search, parallel search, URL extraction, and combined search-extract, with CAPTCHA recovery and no API key required.
Extracts academic PDF content from PubMed (via PMC), supporting abstract mode (~1500 chars) and full text extraction.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@google-surf-mcp search for latest AI research papers"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
English | 한국어
Demo only. Actual searches run headless by default (no visible browser). Set
SURF_HEADLESS=falseto make Chrome visible like in the clip above.
Google search MCP. No API key. Just works.
One MCP replaces three: search + URL fetcher + academic-paper extractor.
✅ Actually works (tested 6 free Google search MCPs, all failed)
✅ Search + URL + academic PDF extract in one MCP (replaces the search MCP + fetch MCP + academic-search MCP combo)
✅ Academic PDFs extracted inline: arxiv, biorxiv, Nature, OpenReview, NeurIPS, JMLR, PMLR, Springer, PubMed (via PMC)
✅ search_extract defaults to abstract mode (~1500 chars/result, token-cheap), mode="full" for whole bodies
✅ Sponsored ads + knowledge panels dropped (geometric verification, not just text matching)
✅ CAPTCHA recovery in 4 modes: OS notification (default) / SURF_HEADLESS=false / SURF_REMOTE_DEBUG / SURF_CLOUD_MODE (fail-fast)
✅ No API key, no proxies, no solver
5 tools: search / search_parallel / extract / search_extract / health
Plug it into any MCP client and you get Google search as a tool.
No CAPTCHA solver. When CAPTCHA fires on any tool, a Chrome window opens for a human to solve. Each solve preserves the profile's reputation with Google.
First call auto-bootstraps the warm profile. Designed for local use. For headless / serverless environments set SURF_CLOUD_MODE=true (fail-fast on CAPTCHA, worker pool disabled).
Related MCP server: Google Search Engine MCP Server
result | |
sequential | ~1.5s/query (first call ~4s, includes setup) |
parallel x4 | ~1.5s wall (first call ~9s, includes pool warm) |
parallel x10 | ~4.5s wall |
search_extract x5 (abstract, default) | ~3s wall |
search_extract x5 (full) | ~5s wall (search + 5 parallel extracts) |
Measured on a workstation with a 1Gb/s connection.
Playwright + persistent Chrome profile
playwright-extra stealth as a cascade fallback tier
Multi-strategy SERP parser + geometric verification (drops sponsored / knowledge_panel / related)
@llamaindex/liteparse for PDF text extraction (PDFium spatial parsing, optional OCR); Mozilla Readability + Turndown for HTML
Resource-blocked images / media / fonts for speed
Auto-bootstrap on first call; pool falls back to single-context after repeated warm failures
Self-healing: runtime parser-strategy reorder (deterministic) + daily cron repair PR (synthesis → optional LLM → triple-gate validation, human review)
Requires Node 18+ and Google Chrome (or Chromium) on the system.
npx google-surf-mcp # actual MCP - register in client configFirst tool call auto-bootstraps the warm profile (you may see Chrome open briefly).
Or local clone:
git clone https://github.com/HarimxChoi/google-surf-mcp
cd google-surf-mcp
npm installIf auto-bootstrap fails (rare), run it manually:
npm run bootstrapOverride paths if needed:
CHROME_PATH=/path/to/chrome SURF_TZ=America/New_York npm run bootstrapPaste this into your ~/.claude.json:
{
"mcpServers": {
"google-surf": {
"command": "npx",
"args": ["-y", "google-surf-mcp"]
}
}
}Restart Claude Code. Done. search, search_parallel, extract, search_extract, health are now available.
For other MCP clients, use the same JSON shape in their config file.
Local clone variant:
{
"mcpServers": {
"google-surf": {
"command": "node",
"args": ["/abs/path/to/google-surf-mcp/build/index.js"]
}
}
}search(query, limit?) - single query, ~1.5s. Returns title / url / snippet. Sponsored ads + knowledge-panel dropped (response includes dropped count + dropped_reasons). Results cached 24h (SURF_CACHE_TTL_SEARCH_MS=0 to bypass).
search_parallel(queries[], limit?) - pool of 4, max 10 queries per call.
extract(url, max_chars?, mode?) - fetch a URL, return article content.
mode="full" (default): whole body. HTML via Readability, PDFs via liteparse (spatial parsing, multi-column reading order).
mode="abstract": ~1500-char survey (PDF page 1 or HTML meta description). Triage relevance before paying for full text.
mode="metadata": PDF page count only.
Response: content, title, excerpt, length, is_pdf, page_count, extraction_quality. Failures return { error }, never throw.
search_extract(query, limit?, max_chars?, mode?) - search + parallel extract in one call. Default mode="abstract" returns SERP enriched with ~1500-char summaries (cheap triage). Use mode="full" when you actually need the article texts (slower, more tokens).
health() - server status. Response: cascade / pool (warmFailures + fallback) / rateLimiter / cache / telemetry / selfHealing (current strategy order + stats) / config. Call it if searches start failing — pool.fallback=true or rising cascade.totalCaptchas are the usual culprits.
var | default | notes |
| auto-detected | absolute path to Chrome binary |
|
| where the warm profile lives |
|
| browser locale |
| system tz | e.g. |
|
| set |
|
| set |
|
| idle ms before closing the sequential ctx and pool. |
|
| set |
|
| default |
|
| OCR scanned/image PDFs via Tesseract (slower; off by default) |
|
| headless/serverless mode: TLS bypass + |
|
| pin a single stealth mode (chosen by |
|
| initial stealth tier — only consulted when |
|
|
|
|
| internal cap on Google-facing requests per minute |
|
| search cache TTL (24h); |
|
| LRU cap per cache namespace |
|
| cache directory |
|
|
|
|
|
|
|
| set |
|
| directory for jsonl telemetry files. UTC-dated one file per day ( |
|
| per-strategy outcome tracking + persisted reordering. Healing must win by 3 outcomes before reorder kicks in, so single-call flapping is impossible. Set |
|
| persistence path for healing state. Atomic tmp+rename writes; debounced 5s. |
|
| opt-in for LLM-assisted selector repair in the workflow-only |
| — | your Anthropic key. Read only when |
CAPTCHA in 4 modes (picked automatically from env):
default (local desktop): OS notification fires, headed Chrome opens, human solves, call retries
SURF_HEADLESS=false: headed Chrome opens, no notification (user is already watching)
SURF_REMOTE_DEBUG=true: DevTools port + instructions printed, attach chrome://inspect locally to solve
SURF_CLOUD_MODE=true: fail-fast with CAPTCHA_REQUIRED error
Headed Chrome opens to a plain search box instead of CAPTCHA: just type any query in the box and press Enter. Subsequent calls work.
"Chrome not found": install Chrome or set CHROME_PATH.
Stale selectors: two-layer mitigation — runtime per-strategy reorder (SURF_SELF_HEALING, deterministic) + daily cron that opens draft PRs with candidate fixes (SURF_LLM_HEAL optional, human review required, never auto-merged).
Searches feel slower than the Numbers table: check health().pool.fallback. true means the worker pool gave up after 3 warm failures and is using a single context. Usually fixed by npm run bootstrap to refresh the seed profile.
SSRF: extract blocks localhost, private IPs, AWS metadata by default. Set SURF_ALLOW_PRIVATE=true to allow them.
See CHANGELOG.md.
MIT
Fetch one public URL -> clean article text. HTML via Mozilla Readability; academic PDFs (arxiv/biorxiv/Nature/OpenReview/NeurIPS/JMLR/PMLR/Springer/PubMed-via-PMC) auto-detected via Content-Type, %PDF magic, citation_pdf_url meta, and per-domain URL rules. Tiered depth: mode="abstract" returns ~1500 chars (PDF page 1 or HTML meta description) -- cheap survey to triage relevance before paying for full body. mode="full" (default) returns the whole article. Best-effort: failures return an errorInfo instead of throwing.
| Name | Required | Description | Default |
|---|---|---|---|
| url | Yes | Public http(s) URL. Loopback/private IPs blocked unless SURF_ALLOW_PRIVATE=true. | |
| mode | No | Extraction depth. `full` = whole article body (default; uses Playwright if needed). `abstract` = cheap survey: PDF page 1 OR HTML meta description (~1500 chars); use to triage relevance before paying for full text. `metadata` = page count only (PDF). Academic PDFs (arxiv/biorxiv/Nature/OpenReview/NeurIPS/JMLR/PMLR/Springer/PubMed-via-PMC) are auto-detected; abstract mode skips Playwright for them. | full |
| max_chars | No | Truncate body to this many chars (default 8000, set via SURF_EXTRACT_MAX_CHARS). |
| Name | Required | Description |
|---|---|---|
| url | No | |
| meta | No | |
| error | No | |
| title | No | |
| is_pdf | No | |
| length | No | |
| content | No | |
| excerpt | No | |
| elapsed_ms | No | |
| page_count | No | |
| extraction_quality | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description goes well beyond the readOnlyHint/idempotentHint annotations by detailing HTML processing via Mozilla Readability, academic PDF auto-detection mechanisms, tiered depth modes, and best-effort errorInfo handling. This gives the agent a comprehensive understanding of the tool's behavior.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise and front-loaded with the main purpose. Every sentence earns its place, covering core functionality, domain-specific handling, tiered modes, and error behavior without redundancy.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's complexity—HTML vs. PDF handling, multiple modes, and error handling—the description covers all essential aspects. The output schema exists, so return-value details are not needed. The description is complete enough for safe and correct invocation.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema already provides complete descriptions for all three parameters (url, mode, max_chars), including explanations of modes and the truncation behavior. The tool description adds no meaningful parameter-specific information beyond what the schema already documents, so the baseline of 3 applies.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool fetches a single public URL and returns clean article text. It distinguishes itself from sibling tools like search or search_extract by emphasizing it operates on a known URL rather than searching for one.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides clear context: the tool accepts a public URL, and advises using abstract mode to triage relevance before full extraction. However, it does not explicitly contrast with the sibling search_extract tool, so no explicit alternatives or exclusions are given.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
MCP server status: cascade mode + transitions, rate-limiter usage, cache size, config. Call this if searches start failing or returning empty -- check cascade.totalCaptchas and rateLimiter.queueSize, and reduce search volume if they are high.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| pool | No | |
| cache | No | |
| error | No | |
| config | No | |
| cascade | No | |
| version | No | |
| telemetry | No | |
| rateLimiter | No | |
| selfHealing | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true and idempotentHint=true, so the safety profile is covered. The description adds valuable context by naming specific fields to inspect (cascade.totalCaptchas, rateLimiter.queueSize), which helps the agent understand what the status output contains and how to interpret it. No contradiction with annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two sentences: the first front-loads the status categories, and the second gives targeted diagnostic trigger and interpretation guidance. Every sentence earns its place with no redundancy or filler.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a zero-parameter, read-only tool with an output schema, the description covers purpose, when to use it, and how to act on results. It is fully sufficient for an agent to select and invoke the tool correctly without needing additional documentation.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool has zero parameters, and the description correctly avoids discussing any. Per the rubric, a zero-parameter tool gets a baseline of 4 because there is no parameter semantics to clarify.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly identifies the tool as a status/health check for the MCP server and lists specific components (cascade mode, rate-limiter, cache, config). It is distinguishable from sibling search/extract tools by focusing on server diagnostics. However, it lacks an explicit verb like 'returns' or 'checks', making it slightly less direct.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explicitly states when to call the tool: 'Call this if searches start failing or returning empty'. It also provides actionable follow-up advice about reducing search volume. However, it does not explicitly mention when not to use it or name alternative tools, so it falls short of a 5.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Single Google search -> title/url/snippet per result. Results are cached 24h, so repeating a query is free -- prefer re-querying over caching results yourself. For latest/today/breaking queries set SURF_CACHE_TTL_SEARCH_MS=0 to bypass the cache. Default limit 10 (max 20). First call ~4s (Chromium warmup), then ~2s. On CAPTCHA a visible Chrome opens for a human to solve (shared-IP protection); SURF_CLOUD_MODE=true makes it fail-fast instead.
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | Max results (default 10). | |
| query | Yes | Google search query. Use site: filters and quotes for exact match. |
| Name | Required | Description |
|---|---|---|
| meta | No | |
| error | No | |
| query | No | |
| results | No | |
| elapsed_ms | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Despite annotations marking readOnlyHint, the description goes far beyond by disclosing caching behavior (24h TTL), timing expectations (~4s first call, ~2s subsequent), CAPTCHA fallback with human interaction, and the fail-fast env var. This provides valuable operational context that annotations alone do not convey.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is somewhat long but front-loaded with the core purpose. Each sentence contributes unique operational facts (performance, caching, CAPTCHA, env vars) without redundancy. The 'Default limit 10 (max 20)' slightly duplicates schema, but overall it earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The description covers all key operational aspects: output structure, caching, performance, error handling (CAPTCHA), and configuration via env vars. With an output schema present, return values are already defined, and the description fills the remaining gaps comprehensively.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema fully documents both parameters ('query' and 'limit') with descriptions and constraints, so the description adds little param-specific detail beyond the default limit and max. The caching and env var info are not directly about parameter syntax, so baseline 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states 'Single Google search -> title/url/snippet per result', specifying the exact resource (Google search) and output. The word 'single' differentiates it from sibling search_parallel, and the context makes it unambiguous what this tool performs.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives clear practical usage guidance: 'prefer re-querying over caching results yourself', how to bypass cache for fresh queries, and the CAPTCHA/cloud-mode behavior. It implies the single-query use case but doesn't explicitly contrast with alternatives like search_parallel, so no explicit exclusion.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
One-shot Google search + parallel extract of the top results. Default mode="abstract" returns SERP enriched with ~1500-char abstracts per result -- a cheap survey of what the top results actually contain, far fewer tokens than fetching all bodies. Switch to mode="full" only when you need the actual article texts (slower, much more tokens). Per-page extract failures are isolated. Disabled in cloud mode.
| Name | Required | Description | Default |
|---|---|---|---|
| mode | No | Extraction depth per result. `abstract` (default) = cheap survey, ~1500 chars/result, ideal for relevance triage. `full` = whole body per result, slower and far more tokens; only when you actually need the article texts. | abstract |
| limit | No | Number of results to extract (default 5, max 10). | |
| query | Yes | Search query. | |
| max_chars | No | Truncate each result body. Default depends on mode: ~1500 for abstract, 8000 for full (SURF_EXTRACT_MAX_CHARS, capped at 20000 here). |
| Name | Required | Description |
|---|---|---|
| meta | No | |
| error | No | |
| query | No | |
| results | No | |
| elapsed_ms | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true and openWorldHint=true. The description adds valuable behavior beyond this: per-page extract failures are isolated, cloud mode disables the tool, and mode affects token consumption. These details help the agent anticipate failure modes and cost, without contradicting any annotation.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is three sentences, front-loaded with the core function, then elaborates on mode selection and constraints. Every sentence provides actionable information with no filler or repetition.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The tool has an output schema, annotations, and a description that covers mode differences, token cost, failure isolation, and cloud limitation. This is complete for a combined search+extract tool with good structured metadata; nothing critical is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, with detailed descriptions for all parameters including mode trade-offs. The description reinforces the mode semantics but does not add significant meaning beyond what the schema already provides, so the baseline of 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description opens with a specific verb+resource: 'One-shot Google search + parallel extract of the top results.' It clearly distinguishes this tool from siblings like search, search_parallel, and extract by combining both functions. The mode parameter is explicitly tied to the purpose, reinforcing clarity.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides explicit guidance on when to use abstract vs full mode, including trade-offs on tokens and speed ('far fewer tokens', 'only when you need the actual article texts'). It also notes that the tool is disabled in cloud mode. It does not name sibling tools as alternatives, but the usage context is clearly implied.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Run 2-10 Google searches concurrently. Use to compare multiple angles in one call. Each query counts against the internal rate limit (~10/min) -- do not loop this for bulk scraping. First call adds 5-10s pool warmup. Per-query failures are isolated in the results array. Disabled in cloud mode.
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | Max results per query. | |
| queries | Yes | 2-10 queries to run concurrently. |
| Name | Required | Description |
|---|---|---|
| meta | No | |
| error | No | |
| results | No | |
| elapsed_ms | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Beyond the readOnlyHint annotation, the description discloses several critical behaviors: internal rate limit (~10/min), first-call warmup (5-10s), per-query failure isolation, and cloud-mode disabling. This significantly exceeds what annotations alone provide.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Four sentences, each carrying essential information. The main action is front-loaded, and there is no filler or repetition of schema details.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The description covers key behavioral constraints, rate limits, failure handling, and environment restrictions. An output schema exists, so return-value explanation isn't necessary. Minor inconsistency between schema minItems (1) and description (2-10) is a schema issue, not a description gap.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100% for both parameters, so the schema already explains 'queries' and 'limit'. The description reinforces the 2-10 query count and adds rate-limit context, but doesn't add new parameter-level meaning beyond the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action: 'Run 2-10 Google searches concurrently.' It identifies the resource (Google searches) and the specific parallel capability, distinguishing it from the sibling 'search' tool.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides a clear use case ('compare multiple angles in one call') and an explicit when-not-to-use ('do not loop this for bulk scraping'). However, it doesn't explicitly name an alternative tool for single searches, so it misses the full 'alternatives' criterion for a 5.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.
v0.1.0extracthealthsearchsearch_extractsearch_parallelEach tool has a clearly distinct purpose: single search, parallel searches, content extraction, combined search+extract, and health status. Although search and search_parallel both involve searching, the parallel variant is explicitly for multiple concurrent queries and the combined tool adds extraction, leaving no ambiguity.
Tool names are all lowercase with underscores and follow a predictable pattern: search and extract are verb-noun style, while search_parallel and search_extract are compound verbs. Health is a slight deviation as a noun, but the overall style is consistent and readable.
With 5 tools, the server is well-scoped for its purpose of Google search and content extraction. Each tool serves a distinct need without redundancy, and the count feels neither sparse nor bloated.
The tool surface covers the full workflow: searching, parallel searching for comparisons, extracting content from URLs, and combining search+extract for efficient surveys. The health tool fills an operational niche. No obvious dead ends or missing essential operations for the server's stated purpose.
Serper MCP — wraps the Serper Google Search API (serper.dev)
Live AI-native web search with citations. One tool for every MCP client. Flat per-request pricing.
MCP server for Google search results via SERP API
Docs: https://docs.keenable.ai/mcp-server Keenable is a free, remote MCP server that gives agents access to the web index. Search the web with ranked results and date/site filters, then fetch any indexed page as clean markdown. Works out of the box with no account or API key.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/tamir-ariunsukh/google-surf-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server