Integration

Olostep integration

Olostep is a web search, scraping and crawling API built for AI agents. Scrape a URL into markdown, HTML, text or structured JSON (parsers or LLM extraction), search the web in natural language, get AI answers backed by live browsing, map every URL of a site, run batch scrapes and crawls, and set up page monitors and scheduled calls. Every scrape, search, answer, map, batch and crawl spends Olostep credits (each tool states its cost). Not exposed: credit top-up purchases (they charge a saved card), file uploads and downloads (pre-signed URLs), hosted file links and screenshots, the monitor planning, run-log and agent-log routes, the partner and AirOps routes, and monitor notification targets and webhook URLs (set those in the Olostep dashboard). Olostep's terms require respecting site terms, robots.txt and rate limits; no API rate limit is published. Authentication is an API key from the Olostep dashboard sent as Authorization: Bearer <key>.

Connect Olostep How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in Olostep.Built from the 34 Olostep tools below, on the connection you add once.

What you can do with Olostep

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.

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 monitorPOSTCreates a monitor that checks pages on a schedule and records changes; it starts in status provisioning and becomes active once planned. Runs may spend credits; the docs state no monitor price.
delete_monitorDelete monitorDELETEDeletes a monitor (status becomes deleted) and stops its scheduled runs.
get_monitorGet monitorGETReturns a monitor with its schedule, status, last run and snapshot total_count. Notification targets and webhook URLs are not returned.
list_monitorsList monitorsGETLists the team's monitors with query, schedule, status and last run. Deleted monitors are left out unless include_deleted is true. Notification targets and webhook URLs are not returned.
pause_monitorPause monitorPOSTPauses an active monitor: its scheduled runs stop and status becomes paused. Only active monitors can be paused; restart with resume_monitor.
resume_monitorResume monitorPOSTResumes a paused monitor: its scheduled runs restart and status becomes active. Only paused monitors can be resumed.
update_monitorUpdate monitorPOSTChanges a monitor's frequency and/or metadata; fields left out are kept. A new frequency recreates the schedule and sets status back to active. Refused with 409 while provisioning.

Schedule, 4 operations

create_scheduleCreate schedulePOSTCreates a schedule that calls an API (an Olostep route or any URL) once at execute_at, or repeatedly by cron or natural-language text. Schedules are free; each run bills the call it makes.
delete_scheduleDelete scheduleDELETEDeletes a schedule and stops all its future runs.
get_scheduleGet scheduleGETReturns one schedule: method, target endpoint, cron expression or one-time run time, timezone and state.
list_schedulesList schedulesGETLists the team's schedules with method, target endpoint, cron expression or run time, timezone and state. Deleted schedules are left out unless include_deleted is true.

Batch, 3 operations

create_batchCreate batchPOSTStarts an async batch scrape of many URLs and returns its id. Poll get_batch until status is completed, then read list_batch_items. Costs 1 credit per URL.
get_batchGet batchGETReturns a batch's status (in_progress or completed), total and completed URL counts and credits consumed. Poll it after create_batch, then read results with list_batch_items.
update_batchUpdate batchPATCHChanges a batch's metadata only: keys sent are added or updated, keys set to an empty string are deleted, other keys are kept.

File, 3 operations

delete_fileDelete fileDELETEPermanently deletes a file and its stored data from Olostep.
get_fileGet fileGETReturns a file's metadata: filename, size in bytes, purpose (context or batch) and status.
list_filesList filesGETLists the team's completed JSON files with id, filename, size and purpose (context or batch).

Answer, 2 operations

create_answerCreate answerPOSTAnswers a task by searching and browsing the web, in 3 to 30 seconds, and returns the answer in result.json_content. Costs 20 credits per request.
get_answerGet answerGETReturns an answer already produced by create_answer, by its id, including result.json_content.

Crawl, 2 operations

create_crawlCreate crawlPOSTStarts an async crawl from a URL through its subpages and returns its id. Poll get_crawl until status is completed (1 to 10 minutes), then list_crawl_pages. Costs 1 credit per page crawled.
get_crawlGet crawlGETReturns a crawl's status (in_progress or completed), settings, pages_count and credits consumed. Poll it after create_crawl, then read pages with list_crawl_pages.

Map, 2 operations

create_mapCreate mapPOSTReturns the URLs found on a website (up to about 120 s on large sites). When a cursor is returned, call again with it for the next set. Costs 1 credit plus 1 per extra 1000 URLs returned.
get_mapGet mapGETReturns a map already produced by create_map, by its id: the URL list, its count and the next-page cursor.

Scrape, 2 operations

create_scrapeCreate scrapePOSTScrapes one URL and returns its content in the requested formats (markdown, html, text, json), page metadata and a retrieve_id. Costs 1 credit by default; parsers cost 1 to 5 credits and llm_extract 10 credits per scrape.
get_scrapeGet scrapeGETReturns a scrape already produced by create_scrape, by its id, with its content and page metadata, during its retention period (about 7 days by default).

Search, 2 operations

create_searchCreate searchPOSTSearches the web for a natural-language query and returns deduplicated links with title and description. Costs 5 credits, plus scrape credits per page when scrape_options is set.
get_searchGet searchGETReturns a search already produced by create_search, by its id, with the links and any scraped content stored at search time. No new scrape and no new charge.

Content, 1 operation

retrieve_contentRetrieve contentGETReturns the content of a scraped, batched or crawled page by its retrieve_id (from create_scrape, list_batch_items or list_crawl_pages). Content over 6 MB is not returned inline (size_exceeded).

Credit, 1 operation

get_credit_infoGet credit infoGETReturns the team's remaining Olostep credits, the per-lot breakdown with expiry dates, the active subscription and whether usage is still allowed. Costs no credits.

Feedback, 1 operation

create_feedbackCreate feedbackPOSTSends feedback to the Olostep team about a wrong, incomplete or failed API result (category and note required). Costs no credits.

Item, 1 operation

list_batch_itemsList batch itemsGETLists a batch's processed URLs with custom_id, url and retrieve_id; pass each retrieve_id to retrieve_content for the page content. Paged: pass back cursor while one is returned.

Monitor event, 1 operation

list_monitor_eventsList monitor eventsGETLists a monitor's snapshot events, newest first, each with run_id, time, whether a change was detected and a summary. Paged with next_cursor; count_only=true returns only total_count.

Page, 1 operation

list_crawl_pagesList crawl pagesGETLists a crawl's pages with url and retrieve_id; pass each retrieve_id to retrieve_content for its content. Paged: start with cursor 0, then pass back the returned cursor.

Recent, 1 operation

list_recent_batchesList recent batchesGETLists the team's recent batches with their id, status, creation time and URL counts.

Use Olostep from Claude Code, Cursor or Codex

Connect Olostep in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its Olostep 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 Olostep

Create your workspace
Sign up free. Nothing to install.
Connect Olostep once
Add your Olostep 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.