Integration

Anakin integration

Anakin is a web data API for agents: scrape any URL to markdown, HTML or AI-extracted JSON (one at a time, inline or as a job, or 10 per batch), run a web search with citations, run a multi-step agentic research job, map a site's URLs, crawl a site, and drive an AI cloud browser through a natural-language task. Wire runs pre-built read and write actions on hundreds of websites (search listings, read a profile, fill a form): discover an action, read its parameters, run it and fetch the job. Website monitors watch a page, a whole site or a Wire action on a schedule and record changes, and AI Visibility asks several AI answer engines the same question. Usage is billed in prepaid credits per call (scrape 1, search 3, agentic search 10 plus 1 per URL, crawl 1 per page, monitors 2 to 3 per check, Wire actions per the catalog). Not exposed: signing in to a Wire site with a password, identity sources (password vaults), saved browser sessions (manage them in the Anakin dashboard; their listing answer is not documented), webhook endpoint management, monitor snapshots, alert deliveries and test alerts, and the WebSocket Browser API. Authentication is an Anakin API key (dashboard, 300 free credits) sent in the X-API-Key header.

Connect Anakin How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in Anakin.Built from the 33 Anakin tools below, on the connection you add once.

What you can do with Anakin

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Each badge shows the default agent permission of an endpoint (allowed: an agent calls it without asking; ask first: you are asked before an agent calls it, when the agent or chat asks permission for sensitive actions; never: no agent may call it), which every user can change endpoint by endpoint in their credential settings.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

Monitor, 7 operations

create_monitorCreate monitorPOSTAgents: ask firstCreates a scheduled monitor that checks a page, a whole site or a Wire action every intervalMinutes (min 15) and records changes, optionally alerting a webhook or emails. Billed per check from then on; starts at once unless isActive false.
delete_monitorDelete monitorDELETEAgents: ask firstPermanently deletes a monitor with all its snapshots and change history. Irreversible.
get_monitorGet monitorGETAgents: ask firstReturns one monitor's full configuration and status: URL, scope, mode, interval, active and expiry state, next and last check, credit cost per run and alert settings. Free.
list_monitorsList monitorsGETAgents: ask firstLists your website monitors with their URL, scope, mode, interval, active and expiry state, next and last check, credit cost per run and alert settings. Free.
pause_monitorPause monitorPOSTAgents: ask firstPauses a monitor's schedule: no more checks (and no more billing) until resume_monitor. Returns the updated monitor with isActive false.
resume_monitorResume monitorPOSTAgents: ask firstResumes a paused or expired monitor with no end date, so checks (billed) start again. Returns the updated monitor with isActive true. 402 monitor_limit_reached past your plan's active-monitor cap.
update_monitorUpdate monitorPUTAgents: ask firstReplaces a monitor's whole configuration (same fields as create_monitor): read it with get_monitor first and send back every setting with your change; a left-out setting reverts to its default. Changes a billed schedule.

Web page, 3 operations

scrape_urlScrape urlPOSTAgents: ask firstScrapes one URL and waits (up to about 90 s) for markdown and, with generateJson, AI-extracted JSON. Billed: 1 credit, +2 for JSON. Check status: failed is answered 200; still running returns an id for get_scrape_result.
scrape_url_jobScrape url jobPOSTAgents: ask firstScrapes one URL as a job (1 credit, +1 summary, +2 JSON); unlike scrape_url it offers formats: links, images, summary, screenshots. Waits for the job and returns the finished result; gives up after 5 minutes.
scrape_urls_batchScrape urls batchPOSTAgents: ask firstScrapes 1 to 10 URLs in parallel as one job; one result per URL. Billed: 1 credit per URL (+2 per URL with generateJson). Waits for the job and returns the finished result; gives up after 5 minutes.

Ai visibility search, 2 operations

get_ai_visibility_searchGet ai visibility searchGETAgents: ask firstReturns an AI Visibility search's status, every engine's answer so far (summary, raw answer, latency, credits, consensus or outlier verdict) and the synthesis. Free; poll until completed or failed.
run_ai_visibility_searchRun ai visibility searchPOSTAgents: ask firstAsks AI answer engines (ChatGPT, Gemini, Google AI Overview) one question and compares the answers. Billed per engine; never cached. Waits for the job and returns the finished result; gives up after 10 minutes.

Browser task, 2 operations

get_browser_task_resultGet browser task resultGETAgents: ask firstReturns a browser task's status (running, completed, failed, timed_out) and result: the agent's output, the steps it took, iterations and duration. Free; poll until it settles.
run_browser_taskRun browser taskPOSTAgents: ask firstRuns a natural-language task in an AI-driven cloud browser; it can ACT on websites (fill and send forms; payments refused). Billed. Waits for the job and returns the finished result; gives up after 6 minutes.

Research job, 2 operations

get_agentic_search_resultGet agentic search resultGETAgents: ask firstReturns the status and report of an agentic research job started by run_agentic_search: summary, structured_data and the schema used. Free; poll until completed or failed.
run_agentic_searchRun agentic searchPOSTAgents: ask firstRuns a multi-stage AI research job (search, scrape sources, extract): a summary plus structured_data. Billed: 10 credits + 1 per URL scraped. Waits for the job and returns the finished result; gives up after 10 minutes.

Site, 2 operations

crawl_siteCrawl sitePOSTAgents: ask firstCrawls a site and returns each page's markdown and HTML. Billed: 1 credit per page, charged upfront on maxPages (default 10, max 100). Waits for the job and returns the finished result; gives up after 10 minutes.
map_siteMap sitePOSTAgents: ask firstDiscovers a site's URLs from a starting URL (internal links, optionally subdomains and external links, keyword filter). Billed: 1 credit. Waits for the job and returns the finished result; gives up after 5 minutes.

Wire action, 2 operations

run_wire_actionRun wire actionPOSTAgents: ask firstRuns a Wire action by action_id: a read extracts data, a write ACTS on the site (submits a form, adds to cart). Billed per the action's credits. If the answer has job_id, poll get_wire_job; else data is inline.
search_wire_actionsSearch wire actionsGETAgents: ask firstFinds Wire actions (pre-built actions on hundreds of websites) for a natural-language intent, ranked: action_id, catalog slug, credit cost, auth needs and required and optional inputs. Free. Then run one with run_wire_action.

Wire action build, 2 operations

get_wire_buildGet wire buildGETAgents: ask firstReturns a Wire build's status (pending or processing while it runs), the published actions with their action_id, what could not be delivered (skipped) and the catalog_slug. Free; poll until it settles.
request_wire_action_buildRequest wire action buildPOSTAgents: ask firstAsks Wire to build a new action for a website from a goal: it generates, tests and publishes a scraper. Billed: credits held at submission (currently 25), refunded if the build fails. Track it with get_wire_build.

Wire catalog, 2 operations

get_wire_catalogGet wire catalogGETAgents: ask firstReturns one Wire catalog (website) with every action: action_id, type (read or write), mode, auth mode, parameter list and credit cost, plus the login fields for password sign-in sites. Free.
list_wire_catalogsList wire catalogsGETAgents: ask firstLists every website Wire supports (public catalogs plus your private actions): slug, name, domain, category, auth needs and action count. Free.

Ai engine, 1 operation

list_ai_visibility_sourcesList ai visibility sourcesGETAgents: ask firstLists the AI answer engines you can query with run_ai_visibility_search, each with its slug (what you pass in sources) and display label. Free.

Crawl job, 1 operation

get_crawl_resultGet crawl resultGETAgents: ask firstReturns the status and per-page results (markdown, HTML or error) of a crawl job started by crawl_site. Free; poll until completed or failed.

Map job, 1 operation

get_map_resultGet map resultGETAgents: ask firstReturns the status and discovered URLs of a map job started by map_site: same-domain links, external links and counts. Free; poll until completed or failed.

Monitor change, 1 operation

list_monitor_changesList monitor changesGETAgents: ask firstLists a monitor's detected changes, newest first (up to 200): when, an AI or plain summary, and a structured diff (changed fields with before and after values, line counts, or a site's added, removed and changed pages). Free.

Monitor check, 1 operation

run_monitor_nowRun monitor nowPOSTAgents: ask firstQueues an immediate check of a monitor now, outside its schedule (the cadence is unchanged). Billed like a normal check. Returns the scrape jobId (page and site monitors) or a message (wire monitors).

Scrape job, 1 operation

get_scrape_resultGet scrape resultGETAgents: ask firstReturns the status and content of a scrape job (from scrape_url when it timed out, scrape_url_job or scrape_urls_batch): markdown, HTML, JSON, links, summary; a batch job lists one result per URL. Free; poll until completed or failed.

Web, 1 operation

search_webSearch webPOSTAgents: ask firstRuns a web search for a query and returns up to 20 results with URL, title, snippet and dates. Synchronous. Billed: 3 credits per call.

Wire identity, 1 operation

list_wire_identitiesList wire identitiesGETAgents: ask firstLists your saved Wire identities (an account on one site) and their credentials: id (the credential_id for run_wire_action), type, status, expiry. Secrets are never returned.

Wire job, 1 operation

get_wire_jobGet wire jobGETAgents: ask firstReturns a Wire job's status and, when completed, the action's data (can be personal data) or file list, credits used and errors. Free; poll about once per second until completed or failed.

Use Anakin from Claude Code, Cursor or Codex

Connect Anakin in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its Anakin operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect Anakin

Create your workspace
Sign up free. Nothing to install.
Connect Anakin once
Add your Anakin connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.