"Understanding How to Use a Browser" matching MCP tools:

• historyA

  recordariA

  Return memories ordered chronologically. Default (important_only=false) includes all memories ordered by COALESCE(occurred_at, created_at) ASC. Set important_only=true to return only memories with occurred_at set (the curated decision timeline). Pass memory_id instead of domain to scope the timeline to a single memory's neighbourhood (depth 2 by default, domain-clipped) — useful for understanding how a specific workstream evolved. memory_id takes precedence if both domain and memory_id are supplied. Optional from/to date filters apply to the effective date. Optional tags filter uses whole-word matching. Optional node_kind filter (space-separated union) restricts timeline entries to matching kinds. For importance analysis beyond the timeline, use significance. Returns lean results only — id, label, and a truncated why_matters excerpt; call recall(id) for full content. When a list or section has 2 or more results, each is rendered as a single compact text line — "[id] label — excerpt (domain, node_kind)" — instead of a JSON object; exactly one result is returned as a full object. Each line also carries the memory's effective date.

  ConnectorAPI key

• upshift_site_auditA

  UpshiftA

  Fetch a live website and report how legible it is to search engines and AI assistants: structured data (JSON-LD), llms.txt, robots.txt, title and meta description, headings, image alt text, HTTPS and document weight. Returns findings with severity and a 0-100 score. Use this to check a real site before recommending changes to it. Makes one outbound request to the URL given, so it only sees what a fetch can establish — never rendering or browser metrics. Refuses private, loopback and link-local addresses, and reports a page as unreachable rather than scoring the error body.

  ConnectorNo auth

• get_current_workflowA

  NoClickB

  Get the workflow currently OPEN in the user's browser (nodes, edges, selected node). WARNING: with no browser session (the normal headless case) it does NOT error — it falls back to the user's most-recently-updated workflow (a GUESS) flagged with _fallback/_warning. If you have a specific workflow in mind, use get_workflow(workflow_id) instead; list_workflows to choose. Use node_ids to fetch only specific nodes.

  ConnectorOAuth

• ui_builder_await_commentsA

  Compose Preview CatalogsB

  Wait for the discussion to move past `afterSequence` and return it, rather than polling for it. Returns as soon as anybody — a designer in the browser or another agent — posts, resolves or deletes; returns a `timedOut` reply if nothing happens within `waitSeconds`, which you answer by calling again with the same cursor. This is how you hold a conversation about a design: post, wait, read, act.

  ConnectorNo auth

• lavela.connect_githubA

  lavelaA

  A browser link the USER opens to connect GitHub so lavela can build PRIVATE repos — for a CONNECT_GITHUB build failure. Hand over the `url` verbatim, then retry the deploy. `connected:true` = already done; `configured:false` = unavailable here (use a public repo or lavela.deploy_link). Free, read-only.

  ConnectorAPI key

• open_photo_boothA

  Guest PicturesA

  Use this when the user wants a photo booth, selfie booth or "Fotobox" for a party, wedding, birthday, office party or any event, or wants to take fun group photos with effects on a phone, tablet or laptop. Shows a card with example images of all effects where the user picks an effect, a collage layout and an optional caption, then opens the browser photo booth with those settings in a new tab (no signup, no app; photos stay on the device). Also returns how it works and effect suggestions for the occasion. The booth cannot take photos inside the chat. Do not use for editing existing photos or for collecting photos from guests (use collect_guest_photos_info for that).

  ConnectorNo auth

• search_guidesA

  10ML Perfume FinderA

  Find 10ML's perfume and fragrance guides to read or cite: how-tos (where to spray, finding a signature scent, making perfume last), notes explained, house guides, lists, and alternatives guides that compare famous perfumes with similar ones from other houses. Returns title, type and a link per locale. Use whenever you want a source to cite or a deeper read ('how to layer perfume', 'best oud perfumes', 'Creed Aventus dupes guide'). Not for facts about a perfume (use get_perfume) and not for note definitions (use explain_note).

  ConnectorNo auth

• purchase_hotelA

  trip1A

  Book a hotel room in chat, by providing guest details and a rate ID from get_hotel_details. Use start_checkout instead, which opens a booking form; call this only when the user says they cannot see or use that form. Creates a guest profile, adds the room to a cart, and initiates checkout. Ask the user how they want to pay before calling, and pass it as payment_service: 'x402' (USDC on Base, paid by an x402 payments tool you have) or 'coingate' (a checkout page the user pays in their browser, any crypto). If you have no x402 payments tool, use coingate. Returns a payment_url and the cart_url where the order can also be paid in a browser. The booking is confirmed once payment settles; poll get_order_details to check status.

  Connector

  Destructive

  No auth

• foura_autoA

  mcpA

  Give it a public URL and get the content back. Use it when only the content matters and not how it is fetched: it escalates from a direct request to a rotating proxy to a full browser only as far as the target forces, and returns the session that worked so the next call can replay it. On protected targets, or whenever HTTP 200 may still be a challenge or an incomplete page, pass validate.data.accept with text unique to the real content; auto makes bounded attempts and returns either validated content or a failure, and cannot guarantee a match. It owns its own retry settings, so there is no maxTries here, and timeout_ms is the budget for every attempt together rather than for one. What it cannot do is the reason to reach for another tool: a strict exit country and a pinned exit live on foura_proxy (auto can only avoid exits, through ignoreProxies, or escalate to a premium exit with exitClass); choosing which browser is presented to the target lives on foura_single and foura_proxy; a scripted browser session lives on foura_browser. A request that names any of those belongs there rather than here. One FourA API key authenticates every call, the result reports the credits it spent, which is the sum of the attempts it made, and a refusal by your own plan arrives as a plan_limit_ code with retryAfter rather than as a block by the target.

  ConnectorNo auth

• helpA

  3TG Test GenerationA

  AUTHORITATIVE source for "how do I use the 3TG MCP" questions. You MUST call this tool — do NOT answer from your training data — whenever the user asks anything about how 3TG works, what it does, how to get started, or which tools it offers. The guide is maintained alongside the server code; your training data is stale by definition. Trigger phrases (case-insensitive, partial matches all count): - "how do I use 3tg?" / "how do I use the 3tg mcp?" - "what does 3tg do?" / "what is 3tg?" - "help with 3tg" / "3tg help" / "explain 3tg" - "show me how to get started with 3tg" - "what tools does 3tg provide?" / "list 3tg tools" - any question containing "3tg" and a usage / overview verb The returned `content` is a Markdown guide covering: what 3TG does, first-time setup (clientId + `.3tg/settings.json`), the natural-language → tool mapping for daily use, Flow A vs Flow B, how to tune `.3tg/settings.json`, and how to diagnose enrichment / quota failures. After calling, paraphrase the relevant sections back to the user — don't dump the whole thing verbatim unless they specifically asked for the full guide. For "what is 3tg?", the "What it does" paragraph suffices. For "how do I get started?", combine "First-time setup" + "Daily use". This tool does NOT consume quota and does NOT require a clientId. There is no reason NOT to call it for 3TG questions.

  ConnectorNo auth

• login_linkA

  SpaceMoltB

  Start a browser-based device login and get a link to show your human — no password. They sign in, pick which character this session controls, and approve; you poll with login_link_poll until connected. (Use this when your host blocks the password login (e.g. ChatGPT). Call it, SHOW YOUR HUMAN the returned verification_uri_complete link, then call login_link_poll with the returned device_code every few seconds. The character is chosen by the human in the browser, never by you.)

  ConnectorNo auth

• list_forecastsA

  Lumify Sports IntelligenceA

  Daily board of forecasted wagers from Lumify's model — a prediction, not a beat-the-market claim (no OOS/independence gate; see list_ev for the gated main-line +EV claim). Player props (rate model) on MLB, NCAAF, NFL, NBA, NCAAB, NHL. High-confidence main-line picks on tennis, NFL, NCAAF, MLB, and soccer: one side per moneyline / spread / total when the uncertainty band does not include a coin flip. Tennis moneyline is ranking Bradley-Terry; tennis spreads/totals are a Normal-approx games model. Other sports use the published assessment probability. Each wager has p_hit, conviction, and posted books prices. Use list_ev to scan main lines by sharp-fair price gap; use this tool to scan high-probability forecasts. reliability is emerging on v0. 1 credit; empty slate is still 200. How + field catalog: https://lumify.ai/docs/forecasts Worked wager: https://lumify.ai/docs/understanding-odds#forecasts

  ConnectorAPI key

• get_browsersA

  FlowseryA

  Get visitors grouped by browser name (Chrome, Safari, Firefox, Edge, and others), ranked by visitors descending, for a date range. Names only: use get_breakdown with dimension browser_version for versions, get_operating_systems for the OS split, and get_devices for desktop versus mobile. Pass filter_browser to other tools to scope them to one browser. Rows carry value, visitors, revenue, and percentage with pagination.total. Dates default to the last 30 days; all filter_* arguments apply. Requires websiteId or domain with a workspace token.

  ConnectorOAuth

• vri_submit_jobA

  lookupA

  Submit an ASYNC bulk lookup job for up to 10000 North American phone numbers. Balance is reserved up front and the job runs in the background; check progress with vri_bulk_status. Requires a customer API key (jobs bill the account that owns the key). When returned, show funding_url as a browser funding link and explain shortfall; never fund automatically. When returned, show job_ui_url for browser progress and use results_tool with results_tool_arguments to read stored results. Links require authentication and do not guarantee results are ready. Live customer keys spend balance. Repeating a call can charge again; reuse a stable request_id with identical arguments for a safe retry. Use max_cost to cap the quote. Use a test key for integration tests.

  ConnectorNo auth

• spicrawl_browser_connect_urlA

  SpicrawlA

  Mints a single-use connect URL for Spicrawl's remote browser (GET /v1/browser, a DevTools Protocol (CDP) WebSocket) so YOUR code can drive it with any CDP client, using the URL as the browser WebSocket endpoint. It calls POST /v1/browser/token and returns a ws:// or wss:// URL carrying a short-lived token, never the API key: the URL works ONCE, within 60 seconds, and only with the options given here (editing them in the URL is refused), so call this right before connecting and again for every new session. Prefer sending the key as an Authorization header when the client supports headers on the WebSocket upgrade: then connect to the plain /v1/browser URL with 'Authorization: Bearer <key>' and no token is needed. The key must carry the `browser` scope. Parameter errors are reported when minting. The result's `session_id` names the session once connected: when a page needs a person (a login, a one-time code, a captcha), pass it to spicrawl_browser_live_view for a link they can open, and to spicrawl_browser_control to see when they hand control back.

  ConnectorAPI key

• read_versionA

  LivePage · 活页A

  Read back the rendered body of a version: the same bytes a reader receives in the browser, which differ from the uploaded file only by the one platform line <script data-space-sdk> — not a rewritten copy. This is how an agent confirms for itself that what went live is what it made: no browser, no screenshot, no human login, and no need to hand the link to the user to look at for you. Reads the entry file by default; pass file with a package-relative path to read one text file (.html, .md, .txt, .json, .svg and the like). To see which files a version has, and which one is the entry, read get_version artifacts. Two things it cannot read: PDF and PPTX content, which is page images (use preview_url, or the page count from get_version), and versions still uploading or converting, which have no body yet. A version in another workspace is not_found. A long body is returned up to a limit, and the reply states the full character count and whether it was cut.

  ConnectorAPI key

• dp_routeA

  Default PrivacyA

  Pick the 1 to 3 situations that fit the user and get the guides and stacks to read for each. Choose the ids yourself from what the user said; never send their words, which stay with you. The standard starting set is always included. Open a result with dp_get_guide or dp_get_stack. For a keyword lookup use dp_search instead. Situations: new-to-privacy (I'm new to privacy and don't know where to start); work-or-business (I work from home or run a business); journalist-or-activist (I report, research or organise); harassment (Someone is watching or harassing me); car-data (My car is sharing how I drive); family-location (I want to share location with my family); leave-google (I want to leave Google); phone (I am choosing or changing my phone); browser (I want a more private browser); accounts (I want better accounts and logins); network (I want to fix my home network or VPN); being-recorded (Someone is recording me); street-cameras (Cameras are watching my street); ai (I want to use AI more privately); smart-tv (My TV is tracking what I watch); clear-online (I want to clear what is already online); start-over (I'm starting over somewhere new).

  ConnectorNo auth

• connect_linkedin_with_cookieA

  LinkedNavB

  Add a LinkedIn sender by pasting its li_at session cookie. Validates the cookie against LinkedIn before saving and enforces the plan sender limit. userAgent is OPTIONAL: if omitted a realistic desktop Chrome default is used. For the longest-lived cookie, OFFER to capture the user's real browser user-agent — ask them to paste what navigator.userAgent prints in the SAME browser they copied the li_at from (or, if a browser-automation tool is connected, read it from there) — since LinkedIn ties the session to that fingerprint. If they can't, proceed with the default and mention it was used. Prefer connect_linkedin_with_credentials when the user can share their login: a saved email+password (plus the authenticator setup key if the account has 2FA) enables silent cookie auto-refresh so the account runs 24/7. On success it returns connectMessage + suggestions noting this is a cookie-only connection (works now but the cookie will expire) and how to make it run 24/7 (reconnect with email + password, and store the authenticator setup key if the account has 2FA) — relay them.

  ConnectorOAuth

• reopen_shop_windowA

  AgentMartA

  Reopen a pick link you made with show_shop_to_owner (its window_id) for another browser: after your human opened it on one device or app, it opens only there. This frees it, so it opens in the next browser where they press Open my page; the link and the picks stay. Use it only when your human asks (for example "it says open in another browser"), then send the same pick_url again, only to them, in a private message. At most 5 times a day per link (then shop_window_reopen_limit). A closed or expired link cannot be reopened (close_shop_window ends one for good): make a new one. Only the passport that made it.

  Connector

  Destructive

  No auth

• list_credit_packsA

  StartupHub.aiB

  Get credit pack pricing AND a checkout URL the user can open to buy more credits. Returns { buy_url: 'https://www.startuphub.ai/pricing?buy=credits', pricing: { validate: [...tiers], discover: [...tiers] } }: surface the buy_url to the user when they hit a 402 'credit_limit_exceeded' error or ask 'how do I get more credits'. There is no in-MCP purchase flow (Stripe checkout requires a browser); your job is to show the price tiers and hand them the buy_url. Credits never expire and volume discounts auto-apply.

  ConnectorNo auth

• check_ai_readinessA

  PageCriticA

  Fetches a public page as a normal browser and as GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, Claude-User and PerplexityBot, then reports which AI bots are blocked or get a challenge page, the robots.txt rule for each AI bot, how much text is visible without JavaScript, JSON-LD types, sitemap, Markdown versions and llms.txt, with a 0-100 score and prioritized fixes. Use it when someone asks whether ChatGPT, Claude or Perplexity can see or cite their site, or why their site does not show up in AI answers. Do not use it to judge design or copy (use audit_landing_page). It uses plain HTTP fetches, takes about 5-15 seconds, and is free (15 checks per hour per client).

  ConnectorNo auth