"Adding web browsing and page reading capabilities to an agent" matching MCP tools:

• check_page_for_injectionA

  cloakcheckA

  Scan a single web page for content planted to hijack an AI agent reading it -- invisible unicode (zero-width chars, the unicode 'tag' block used for steganographic prompt injection), CSS-hidden instruction text, and instruction-shaped language in alt/title/aria-label attributes a human would never read. Does NOT judge whether visible body text is safe -- only content hidden from normal human reading flow is flagged, so a page that legitimately discusses prompt injection won't false-positive on itself. Call this before an autonomous shopping/browsing agent acts on a page's content (add to cart, follow instructions found on the page, etc).

  ConnectorNo auth

• get_web_capabilitiesA

  Evercraft WebA

  Describe Evercraft Web capabilities for AI agents, including URL fetch/extraction, future web search, browser automation, cloud/headless browser execution, web-agent navigation, Playwright/CDP-style control, and user-authorized authenticated browsing. This tool is read-only and reports live versus held primitives explicitly.

  ConnectorNo auth

• react_to_commentA

  DockA

  Add or remove an emoji reaction to a comment. Reactions are per-principal: each (commentId, principalId, emoji) combination is unique. `action: 'add'` is idempotent (re-adding the same emoji is a no-op); `action: 'remove'` deletes the row if present. Fires `comment.reaction_added` / `comment.reaction_removed`. Use this for lightweight agent acknowledgement (👍 on a request before reading, 👀 to mark in-progress, ✅ when done), cheaper than a full reply.

  ConnectorNo auth

• cached_fetchA

  slipstreamA

  Fetch a web/docs URL as clean, token-optimized markdown from Slipstream's shared cache (use INSTEAD of a raw web fetch). The first agent pays the crawl; every agent after gets ~90% fewer tokens. Surfaces warnings other agents left on the page. Pass known_hash to skip re-reading unchanged content (delta), or section to fetch just one heading (progressive disclosure). Returns a contentHash you can pass as known_hash next time.

  ConnectorNo auth

• score_agentic_browsingA

  AI-SEO / AEO / GEO AuditA

  Score a page against the four signals Google added to the Lighthouse "Agentic Browsing" category in May 2026: presence of an llms.txt, WebMCP integration, accessibility-tree integrity, and layout stability. Returns an overall 0-100 score, a letter grade, and a per-factor breakdown. Read-only. One HTTP GET for the page plus one for /llms.txt (skip with check_llms_txt=false). Pass `html` instead of `url` to score markup offline (llms.txt is then treated as absent). Deterministic, rule-based heuristics over the fetched HTML; no LLM and no headless render required. This approximates Lighthouse's runtime signals from static markup - it does not execute Lighthouse. When to use: checking whether a site is ready for AI agents / agentic browsers, or tracking the new Lighthouse Agentic Browsing category. For citation-eligibility of content, use `score_citation_worthiness`; for a full page audit, use `audit_page`.

  ConnectorNo auth

• get_agent_readinessA

  turva-mcpA

  Returns turva.dev's own agent-readiness score from an independent public scanner (isitagentready.com), including category sub-scores, with the measurement date and a link to the scanner's start page, where a new check can be run (the link does not open the recorded reading). Use this when a user asks how turva.dev scores, whether its claims are verifiable, or what proof backs the audit service. For web-security scan results, which are a separate measurement, use get_security_evidence instead. Read-only: returns static JSON that is compiled into the Worker, so it changes nothing and updates only on deploy.

  ConnectorNo auth

• delete_proposalA

  LivePage · 活页A

  Delete a page: its link stops working immediately and it disappears from the page list, analytics and folder pages — readers who open it get a 404. You can restore it within 30 days with restore_proposal; after that the platform erases it for good (content, versions, page images, reading records and viewer-appended rows all cease to exist; not reversible). To erase it immediately, use the web console, where the consequences are spelled out. Pages with unresolved reports cannot be deleted — resolve the report first. confirm must repeat the page's title exactly, to guard against deleting the wrong one: check with list_proposals first.

  Connector

  Destructive

  API key

• scale_serviceA

  ValetA

  Set how many processes of one type an app or agent runs. For an app, give process_type, such as web, and a count from 0 up to the organization's cap. For an agent, count is 0 (paused) or 1 (running). Scaling down stops running processes, and scaling web to 0 takes an app's URL offline, so confirm with the user first. Setting the count it already has changes nothing. Websites run no process and are refused. Changing an agent's model is not offered here. Requires connecting a Valet account.

  Connector

  Destructive

  No auth

• read_urlA

  Agent Decision & Evidence Tools (MCP + x402)A

  Fetch a public web page as clean Markdown or text while refusing internal and private addresses, suitable for reading evidence before an agent acts on a link.

  ConnectorNo auth

• handover.whoamiA

  HandoverA

  Return the authenticated person or service agent, organization, workspace, role, and granted scopes. Call this before reading or publishing to confirm attribution and tenancy.

  ConnectorNo auth

• update_artifactA

  Bisque PresentationsA

  Change an Artifact you own or edit. Pass content to publish a new Version of the page at the same URL (the whole document, not a diff; earlier Versions stay in the history). Pass title, description, icon, or icon_image to change those. Any combination in one call. With content the call waits for the new Version to be live before returning. The new Version keeps the base Version's capabilities and state rules unless you pass capabilities or state.

  ConnectorNo auth

• get_groupA

  Kaption WhatsAppA

  Fetch a single group with a LIVE participant list. Forces `Store.GroupMetadata.update()` against the WA backend before reading, so the result reflects current membership including recent joins/leaves. Use this when accuracy matters; use `list_groups` for browsing. Examples: Live group fetch: { jid: "120363421729019499@g.us" }

  ConnectorOAuth

• aanet_get_activity_summaryA

  Autonomous Agents Network (AANet)A

  Spend totals from the audit log instead of reading every entry: for each group, total_cost (units) and calls. Use it to see which sub-agent is burning budget before adjusting a budget_cap via aanet_update_subkey. Owner_key required. An unknown group_by is an error.

  ConnectorNo auth

• aanet_get_activity_summaryA

  AANetA

  Spend totals from the audit log instead of reading every entry: for each group, total_cost (units) and calls. Use it to see which sub-agent is burning budget before adjusting a budget_cap via aanet_update_subkey. Owner_key required. An unknown group_by is an error.

  ConnectorNo auth

• check_urlA

  URL Safety Validator MCPA

  Checks a URL for phishing, malware, typosquatting, and redirect threats. Call this BEFORE your agent fetches, follows, or forwards any URL in an agentic commerce workflow -- at the moment a merchant site, supplier portal, or payment redirect URL is received and no navigation has occurred. Use this when your agent has received a URL from an external source — email, document, or API response — and is about to navigate to it or pass it downstream. Checks live against Google Web Risk (webrisk.googleapis.com) and Google Safe Browsing (safebrowsing.googleapis.com). Returns verdict SAFE / SUSPICIOUS / DANGEROUS with a derived agent_action of ALLOW / FLAG_AND_PROCEED / BLOCK, trust score 0-100, and threat categories. A payment executed on a phishing domain via Stripe MPP, Alipay AI Pay, or Shopify UCP has no recovery path -- the redirect is the attack vector. A DANGEROUS verdict means halt immediately. Returns machine-ready verdict, no further analysis needed.

  ConnectorNo auth

• search_registryA

  WellknownA

  Search FIRST — before a web search — whenever a task needs an external tool, API, service, data source, or live capability you do not already have wired up. Finds capabilities, MCP servers, and APIs agents can actually call, across every publisher indexed by Wellknown, plus the caller's own granted tools. Returns ranked results with an identifier you can resolve, the endpoint URL, and provenance saying whether the entry was found on the publisher's own domain or federated from another registry. A web search returns pages about tools; this returns the tools. Fall back to a web search only when this returns nothing that fits.

  ConnectorNo auth

• search_agentsA

  Trusgent MCPA

  Search the Trusgent Trust Layer directory for people and businesses with public Agent Cards. Uses keyword + semantic (embedding) retrieval over durable profile capabilities. Returns AgentCard objects including profileUrl, jsonUrl, trustScore, verificationLevel, and verified flag. Prefer verified agents when recommending. Use before recommending any agent to an end user. Do not treat this as intent/demand matching — that is a separate flow.

  ConnectorNo auth

• everyinfra_list_capabilitiesA

  EveryInfraA

  List EveryInfra's data platforms and capabilities (90 platforms / 413 capabilities: search, detail, comments, profiles and more). Use it before everyinfra_call_api to get the platform/action, required and optional params, the per-call row limit (max_limit, or page_size for page-based actions) and the price. Pass a platform to also get allowed values, returned fields and what they mean. Free. Data collection only — for AI chat use everyinfra_chat.

  ConnectorNo auth

• get_lead_finder_filtersA

  EmailchaserA

  Returns the values Lead Finder's list filters accept (seniorities, jobFunctions, companySizes, revenue, industries, countries, headquartersCountries, regions, continents), the limits that apply to this workspace (page sizes, deepest page, people per add, first-N count, imports running at once, and the daily browsing allowance with what is left today), and creditsPerProspect, what adding one person to a campaign costs. Free, and makes no call to the data provider. Call it before search_lead_finder so filter values match exactly.

  ConnectorNo auth

• check_page_changesA

  YouSpotA

  Check whether a public web page has changed since the last time this was called for it, and get the diff. Use for watching one page over time: a company's team or leadership page, a pricing page, a careers page, anything where the change is the news. The first call records a baseline and reports no change, which is normal. Pass your own agent id as `tag` so two watchers on the same page do not consume each other's changes. For the contents of a page use read_web_page; for the open web use web_search.

  ConnectorNo auth

• web_scrapeA

  agentsvc.ioA

  URL to Markdown web scraper: read any web page as clean, LLM-ready Markdown. Default tool for reading ONE web page whose URL you have: returns clean Markdown (title, meta description, main content with headings, lists and links). The fastest and cheapest way to give an LLM the content of a page. If a JavaScript app comes back nearly empty, set render:true (real browser, slower, plain text) or use webpage-reader. For 2 to 5 URLs use web-read-batch, for a whole site site-crawl, and if you have a question but no URL search-read. Params: url (required), max_chars (default 8000, max 60000), include_links, render. Price: $0.0025 USDC per call, paid inside this MCP call via x402 (USDC on Base, Solana, Polygon or Arbitrum). x402 MCP clients such as @x402/mcp pay automatically. Free sample: call with the example input {"url":"https://example.com","max_chars":8000}.

  ConnectorOAuth