Skip to main content
Glama
800,884 tools. Updated 2026-11-18 17:13

"A server for deep research using Google search" matching MCP tools:

  • AI Overview (AI Mode) scraping via Google Search. Returns AI-summarized key points for a single keyword, ideal for deep research, technical Q&A, long-tail product selection, and cross-border consumer preference analysis using the latest web information. Single-round only; follow-ups require the agent to summarize context and issue a new request. Triggered by: Google AI, AI Overview, AI Mode, Google AI search, AI search, deep research, consumer preference analysis, web summary, long-tail product research, cross-border market insights.
    ConnectorOAuth
  • AI Overview (AI Mode) scraping via Google Search. Returns AI-summarized key points for a single keyword, ideal for deep research, technical Q&A, long-tail product selection, and cross-border consumer preference analysis using the latest web information. Single-round only; follow-ups require the agent to summarize context and issue a new request. Triggered by: Google AI, AI Overview, AI Mode, Google AI search, AI search, deep research, consumer preference analysis, web summary, long-tail product research, cross-border market insights.
    ConnectorOAuth
  • Search the user's files by filename and return matching documents in the deep-research result shape. ALIAS: this is the SAME search as search_files (same data, same permissions) - use it when your client requires the id/title/url search contract (ChatGPT deep research); otherwise prefer search_files for richer file metadata. Each result's id can be passed to fetch (or get_file) to read that document. Read-only; nothing is written, so it is safe to call.
    ConnectorOAuth
  • Create a complete Google Ads Demand Gen campaign (YouTube, Discover, Gmail, Maps and the Google Display Network) in ONE atomic request: budget, campaign (ALWAYS created PAUSED — activate later with adsap_google_set_entity_status), a typeless ad group with channel controls, ad-group-level location/language targeting, and ONE ad. Use dry_run: true first — Google validates the whole chain without creating anything; show the plan, get confirmation, then execute. Two ad types (ad.ad_type): 'multi_asset' = image ad (1-5 headlines <=30 chars, 1-5 descriptions <=90, business name <=25, 1-5 logos, at least one landscape 1.91:1 or square 1:1 image from adsap_google_upload_asset, optional portrait 4:5 and tall-portrait 9:16, combined images <=20); 'video_responsive' = YouTube video ad (1-5 YouTube video ids, 1-5 headlines <=40, 1-5 long headlines <=90 REQUIRED, 1-5 descriptions <=90, 1-5 logos). Daily budget minimum is 5 (Google enforces a 5 USD local-currency-equivalent floor for Demand Gen). Channels: default ALL_CHANNELS; or channels.selected booleans (youtube_in_stream/youtube_in_feed/youtube_shorts/discover/gmail/display/maps — display: true is how you reach the Google Display Network). bidding_strategy 'auto' picks maximize_conversions when the account has conversion tracking, else maximize clicks; target_roas needs conversion history (Google rejects otherwise). ad_account_id is the bare 10-digit customer ID. For Search use adsap_google_create_search_campaign, for Performance Max adsap_google_create_pmax_campaign; for Meta use adsap_create_campaign.
    ConnectorAPI key
  • Adds one check to a finished research and returns the id of the new child research. Use it when the scan skipped a stage, when a person asks about one phrase in particular, or to rebuild the report after new checks. A niche scan already runs the usual checks by itself, so this tool is for what is missing. Checks added this way are free: they do not spend researches of the plan. Checks of a demand research (`kind`): - `competitors`: who holds the search results for a phrase of the research; pass the phrase in `phrase`. - `dynamics`: two years of demand history. - `appstore`: App Store apps for the phrase with ratings and freshness. - `wiki`, `radar`, `youtube`: public interest and community signals. - `paid`: paid search for the phrase and its strongest candidates: volume, advertiser competition, cost per click. - `insight`: a summary of the findings written by the language model. - `report`: the final niche report, written anew from the current data; needs a finished niche scan. Checks of a child research: - `fingerprint`: product sites found by a competitors check, with pricing and app links; the parent is that competitors check. - `reviews`: recent App Store reviews of the leaders; the parent is an App Store research. - `insight` also accepts a competitors check as the parent. The parent must be completed and belong to the account. The check runs in the background: poll get_research with the returned `id` every 20-30 seconds until `status` is `completed`. `insight` and `report` are rebuilt at most once an hour.
    ConnectorNo auth
  • Google search volume, CPC, competition, keyword difficulty and intent for up to 20 comma-separated keywords in one $0.04 GET (Google Ads data). Use when you need monthly search volume for a few keywords in a country. volume = avg monthly searches, cpc in USD, competition 0-1, difficulty 0-100. Up to 100 keywords for the same $0.04: POST /v1/keyword-overview. Keywords without data are listed in missing (still charged). Bad input is free; upstream errors are not charged. Price: $0.04 per successful call, paid over x402 (USDC on Base); pass the payment in _meta["x402/payment"].
    ConnectorNo auth

Matching MCP Servers

  • A
    license
    Not graded
    quality
    D
    maintenance
    Enables deep research tasks using a multi-agent architecture that integrates any LLM and MCP tools. Available via MCP stdio, streamable HTTP, and SSE transports.
    16
    MIT
  • A
    license
    A
    quality
    B
    maintenance
    It enables Claude or other MCP clients to search 17 academic, open-data, archival, independent-web, and Tor dark-web sources in parallel, read HTML, PDF, and .onion pages, recover Wayback snapshots, and filter every query and result through permanent safety checks.
    4
    MIT

Matching MCP Connectors

  • Autonomous deep research reports merging PSFK trend graphs with citable sources.

  • Autonomous buy-side research: diligence, earnings, SEC filings, comp sets. Source-cited real data.

  • Use this when a deep research run needs to look up digital tools and products on uneed.best. Same catalog and same relevance ordering as search_products, returned as `{id, title, url}` documents; pass a result's id to `fetch` for the full profile. Prefer search_products when you want structured product fields directly.
    ConnectorNo auth
  • Attach real search data to one article idea that is already in the project — runs one live Google search per target keyword and returns volume bands and the pages currently ranking for each. Use after create_article_suggestion_with_input, or on any existing suggestion the operator wants judged on data instead of instinct. Opportunity score and cluster placement are returned only when the project has a completed research run to compare against; on a project without one they are absent, and start_research_run is what produces them. Do not use to find new ideas or map a niche — that is start_research_run. Consumes no credits; daily- and monthly-capped per account.
    ConnectorNo auth
  • Open-ended discovery across SMI deep reports, research bundles, and exports when you do not already have a record id — finds candidates by subject name, url, or keyword and returns typed references to pass to fetch. If you already have a record id from the conversation rather than from a search result, call the specific tool instead: smi_get_deep_report or smi_get_export to read one record, smi_check_order_complete for order readiness, smi_list_deep_reports or smi_list_exports for filtered listings, smi_get_report_findings for findings within a report. Deep reports and research bundles are filtered server-side by subject first/last name, broadening to an unfiltered fetch ranked client-side (across id, name, url, summary) when the filtered fetch finds nothing; exports match on url.
    ConnectorOAuth
  • Draft a public reply to one review, using its id from my_reviews. Replaces any existing reply. Two or three sentences in the owner's voice; never offer discounts, argue, or share personal details. Creates a DRAFT only — nothing on Google changes. Show the owner the draft this returns, word for word, and publish it with publish_change only after they say yes.
    ConnectorNo auth
  • Draft deleting one photo from the listing, using its id from my_photos. This one cannot be undone. Creates a DRAFT only — nothing on Google changes. Show the owner the draft this returns, word for word, and publish it with publish_change only after they say yes.
    ConnectorNo auth
  • The top organic results and SERP features Google shows for one query. Use when: "what ranks for X", "who is on page one for X", "is there a featured snippet or AI overview for X". Not for: where one site ranks for a keyword, use seo_get_position (it scans the top 100 for you). Keyword ideas, use seo_search_keywords. The businesses Google Maps ranks for a local search, use maps_search_places. Returns: Record. features[] (featured_snippet, people_also_ask, video, ai_overview and others), results[] (rank, organic_rank, type, url, domain, title, snippet, published_at, rating), related_searches[], ai_overview (text, references[]) or null, results_count, check_url. depth counts organic rows: the default returns 10 organic results whenever Google has them, with the features ranked among them on top. Each People Also Ask question and each forum thread is its own row at its block's rank; for a question, title is the question and snippet the answer, url null when Google's answer is an AI overview. domain is null on a row whose url is still a google.com/goto redirect. location localises the local pack, not the organic rows of a bare service search: put the town in the keyword ("dentist columbus ohio") for local organic results. Paid ads seldom come back, so the page says nothing about who bids on the query; a named rival's search ads are in ads_get_advertiser_ads with platform google. Cost: 1 credit per 10 results, +1 with ai_overview. Default top 10 = 1 credit. Cached 24 hours. Example: {"keyword": "hubspot alternative", "depth": 10, "ai_overview": true}
    ConnectorAPI key
  • The advertisers matching a brand name in Google's ad library, with their ids. Use when: before ads_get_advertiser_ads on Google, because that library is keyed by advertiser id and one brand has one entry per region. Not for: the ads themselves, use ads_get_advertiser_ads with the id this returns. The other libraries, which search their ads directly with ads_search_ads. Returns: List. rows[]: platform, id, name, region, ads_estimate, website. Cost: 1 credit. Cached 7 days. Example: {"platform": "google", "query": "lululemon", "region": "US"}
    ConnectorAPI key
  • Search stored daily top-chart snapshot records from the iOS App Store and Google Play; omitting `date` selects the latest snapshot. Costs 1 credit per record returned (minimum 2), so keep `page_size` to what you need. Filter by app, store, platform, chart, country, or category; use datasets_apps_search for app records or datasets_apps_reviews_search for stored reviews.
    ConnectorNo auth
  • Search this Cubite site's courses by keyword and get each match's public page URL, for linking to or citing it. This is the search half of the search/fetch pair that ChatGPT connectors and deep research require. The whole query is matched as one case-insensitive phrase, so use one or two distinctive words. Courses match on name or description and are listed in name order; a course without a URL slug has no public page and is never returned, while courses hidden from the catalog are. Returns up to 10 results as {id, title, url}; pass a result's id to `fetch` for its citable text. It is not an inventory: to list, count or manage courses use course_list, which has no keyword filter but returns every course, with or without a slug, with its id, settings and enrollment count.
    ConnectorAPI key
  • Search stored daily top-chart snapshot records from the iOS App Store and Google Play; omitting `date` selects the latest snapshot. Costs 1 credit per record returned (minimum 2), so keep `page_size` to what you need. Filter by app, store, platform, chart, country, or category; use datasets_apps_search for app records or datasets_apps_reviews_search for stored reviews.
    ConnectorNo auth
  • The Google Analytics 4 properties the workspace's connected Google account can read, with the account each sits under. The first analytics call; the other analytics tools take a property from here. Use when: "what's in my Google Analytics", "which GA4 property is example.com", before any analytics_get_* call. Not for: Search Console sites, use console_list_properties. Traffic to a site the user does not own, use seo_get_traffic_estimates. Returns: List. rows[]: property (the id to pass on: properties/123456789), property_name, property_type (PROPERTY_TYPE_ORDINARY, SUBPROPERTY or ROLLUP), account (accounts/123), account_name. NotConnected when no Google account is connected, or when the connected one has not granted Analytics; its connect_url is where the user connects or reconnects. Cost: 0 credits. The user's own data is free; it shares the console's 60 calls per minute per workspace. Cached 1 hour. Example: {}
    ConnectorAPI key
  • Starts an AI draft of ad groups for an existing Search campaign and returns draft_id at once; it changes nothing in Google Ads. ADM reads the landing page, groups the keywords by theme and writes responsive search ads per group. Poll get_ad_group_draft every poll_after_seconds (usually done in 1-3 minutes). The finished ad_groups can be passed verbatim to add_ad_groups (preview first). Needs a read_write key and a paid plan. Each call runs several AI calls: do not repeat it for the same input while a draft is pending or running.
    ConnectorAPI key
  • Monthly Amazon search volume for up to 1,000 keywords you already have, in the order given. Use when: "how many people search Amazon for X", comparing Amazon demand with Google demand, sizing a product idea on Amazon. Not for: Google volume, use seo_get_keyword_metrics. Finding keywords, use amazon_search_keywords. Returns: List. rows[]: keyword, volume (monthly Amazon searches; null = no data). Cost: 5 + 5 per 100 keywords. 10 keywords = 6 credits. Cached 7 days. Example: {"keywords": ["standing desk", "desk converter"]}
    ConnectorAPI key
  • Creates a new Google Ads campaign with a budget. The campaign is always created paused: it shows no ads and spends nothing until it is switched on in a separate step. SEARCH campaigns are created with Google search on, Search partners off, the Display Network off, and location targeting set to PRESENCE (people in the targeted locations). Pass search_partners, display_network or geo_target_type to choose other values. On a campaign of any other channel type, each of these three settings is set only when its argument is passed. Search campaigns have no language setting: Google Ads removed it in late September 2026 and matches Search ads to searchers by the language the ads are written in, so write the ads in the language customers search in. Searches in another language that cost money are stopped with add_negative_keywords. languages sets a DISPLAY campaign's language setting in the same request as the campaign; without it the campaign has no language setting, may show to people who use any language, and the result says so. languages is refused on SEARCH. SHOPPING is turned off until Shopping campaign creation is complete: a Shopping campaign needs a Merchant Center account, a feed label and a product group, which this tool does not set yet, so it is refused before anything is sent. VIDEO is refused before anything is sent: the Google Ads API cannot create or change Video campaigns. MULTI_CHANNEL (App) is refused before anything is sent: an App campaign needs an app ID, an app store and a bidding goal, which this tool does not set. The budget and the campaign are created together in one request, so if Google refuses the campaign no budget is left behind. The budget belongs to this campaign alone (it is not a shared budget). A channel type or bidding strategy outside the lists below, a strategy the channel type does not take, or a target missing or given to the wrong strategy, is refused before anything is sent. Bidding strategies by channel type: SEARCH: MANUAL_CPC, TARGET_SPEND, MAXIMIZE_CONVERSIONS, TARGET_CPA, MAXIMIZE_CONVERSION_VALUE, TARGET_ROAS. DISPLAY: MANUAL_CPC, TARGET_SPEND, MAXIMIZE_CONVERSIONS, TARGET_CPA, TARGET_ROAS. Google does not offer MAXIMIZE_CONVERSION_VALUE without a target ROAS on Display campaigns. Args: customer_id: The Google Ads customer ID (numbers only, no hyphens). name: The name for the new campaign. budget_amount_micros: Daily budget in micros of the account's currency (10000000 = 10.00: ₹10 on an INR account, $10 on a USD one). advertising_channel_type: Channel type. One of: SEARCH, DISPLAY. Default: SEARCH. SHOPPING is turned off until Shopping campaign creation is complete; VIDEO is refused (the Google Ads API cannot create Video campaigns); MULTI_CHANNEL (App) is refused (it needs an app ID, an app store and a bidding goal this tool does not set). For Performance Max use create_performance_max_campaign. status: Initial campaign status: PAUSED, the only status a build takes. Every campaign is built paused and is switched on in a separate step; ENABLED is refused, and nothing in that call runs. Default: PAUSED. bidding_strategy_type: Bidding strategy. One of: MANUAL_CPC, MAXIMIZE_CONVERSIONS, MAXIMIZE_CONVERSION_VALUE, TARGET_CPA, TARGET_ROAS, TARGET_SPEND, as the channel type takes it (list above). Default: MANUAL_CPC. target_cpa_micros: Target CPA in micros of the account's currency. Required with TARGET_CPA, refused with any other strategy. target_roas: Target ROAS (e.g., 2.0 for 200%). Required with TARGET_ROAS, refused with any other strategy. cpc_bid_ceiling_micros: Maximize Clicks' max CPC cap — the most Google may bid for one click — in micros of the account's currency (10000000 = 10.00: ₹10 on an INR account, $10 on a USD one). Only with bidding_strategy_type TARGET_SPEND. Optional. start_date: Campaign start date in YYYY-MM-DD format. Optional. end_date: Campaign end date in YYYY-MM-DD format. Optional. search_partners: Whether ads may serve on the Google Search Partners network. SEARCH default: False. Optional. display_network: Whether ads may serve on the Google Display Network. SEARCH default: False. Optional. geo_target_type: Positive location targeting. One of: PRESENCE (only people in the targeted locations), PRESENCE_OR_INTEREST (also people elsewhere who show interest in them), SEARCH_INTEREST. SEARCH default: PRESENCE. Optional. languages: Google Ads language codes (e.g. en, hi, es, pt, de, fr, ja, zh_CN) a DISPLAY campaign may show to. ["all"] sets no language setting on purpose. Refused on SEARCH. Optional. over_comfort_ack: The user's own words or reason when they asked for a budget above their saved comfort (e.g. "user asked for 6,000/day on 2026-09-23"). Recorded with the change by the hosted connector; never sent to Google. Optional. Returns: Dictionary with created campaign and budget resource names; on a DISPLAY campaign also the language codes set (an empty list means every language). Tools revision: 7a2c68c7.
    ConnectorOAuth
  • Creates a Performance Max campaign with a budget, business name, and logo. The campaign is always created paused: it shows no ads and spends nothing until it is switched on in a separate step. PMax campaigns require a business name and logo asset linked at the campaign level (Brand Guidelines). Create these assets first using create_text_asset (for business name) and create_image_asset (for logo), then pass their resource names here. After creating the campaign, create more assets and an asset group using create_text_asset, create_image_asset, create_youtube_video_asset, and then create_asset_group. languages sets the campaign's language setting in the same request as the campaign. Google uses it on YouTube, Display, Discover and Gmail; on Google Search, Performance Max ads are matched by the language the ads are written in. Without languages the campaign has no language setting and may show to people who use any language on those channels; the result then says so. Args: customer_id: The Google Ads customer ID (numbers only, no hyphens). name: The name for the new campaign. budget_amount_micros: Daily budget in micros of the account's currency (10000000 = 10.00: ₹10 on an INR account, $10 on a USD one). business_name_asset_resource_name: Resource name of a text asset for the business name. logo_asset_resource_name: Resource name of an image asset for the logo (square, 1200x1200). status: Initial campaign status: PAUSED, the only status a build takes. Every campaign is built paused and is switched on in a separate step; ENABLED is refused, and nothing in that call runs. Default: PAUSED. bidding_strategy_type: One of: MAXIMIZE_CONVERSIONS, MAXIMIZE_CONVERSION_VALUE. Default: MAXIMIZE_CONVERSIONS. target_cpa_micros: Target CPA in micros of the account's currency, optional with MAXIMIZE_CONVERSIONS. target_roas: Target ROAS (e.g., 2.0 for 200%), optional with MAXIMIZE_CONVERSION_VALUE. final_url_expansion_opt_out: Set True to opt out of final URL expansion. Default: False. start_date: Campaign start date in YYYY-MM-DD format. Optional. end_date: Campaign end date in YYYY-MM-DD format. Optional. languages: Google Ads language codes (e.g. en, hi, es, pt, de, fr, ja, zh_CN) the campaign may show to on YouTube, Display, Discover and Gmail. Not used on Google Search. Optional. over_comfort_ack: The user's own words or reason when they asked for a budget above their saved comfort (e.g. "user asked for 6,000/day on 2026-09-23"). Recorded with the change by the hosted connector; never sent to Google. Optional. Returns: Dictionary with created campaign and budget resource names, and the language codes set (an empty list means every language). Tools revision: 7a2c68c7.
    ConnectorOAuth