---
title: "tool si.act.crawl"
description: "a `ScrapeRequest` whose scope names the site's seeds, with sitemap-driven discovery on: the `ScrapeRun`, queued, or the run that already stands for the same template version and scope inside its freshness; every page lands as a capture through the one pipeline, and each page that could not be read is a gap with its…"
section: reference
address: /reference/tools/si-act-crawl
contract: 0476e35e6e275db5
---

# tool si.act.crawl

a `ScrapeRequest` whose scope names the site's seeds, with sitemap-driven discovery on: the `ScrapeRun`, queued, or the run that already stands for the same template version and scope inside its freshness; every page lands as a capture through the one pipeline, and each page that could not be read is a gap with its…

[figure: Over MCP, si.act.crawl takes ScrapeRequest and returns a card whose body is ScrapeRun.]

the `ScrapeRun`, queued, or the run that already stands for the same template version and scope inside its freshness; every page lands as a capture through the one pipeline, and each page that could not be read is a gap with its reason. The call needs the act grant at the brand rung.



| | |
|---|---|
| input | [ScrapeRequest](/concepts/scraperequest) |
| result | a card whose body is [ScrapeRun](/concepts/scraperun) |
| read-only | no |
| destructive | yes |
| card address | `ui://scale-intelligence/cards/ScrapeRun` |
| HTTP route | [`/v1/act/crawl`](/reference/act/crawl) |





| field | type | required | note |
|---|---|---|---|
| `template` | `id` | no | an extraction template; any of its parts may be overridden below |
| `scope` | [CrawlScope](/concepts/crawlscope) | no | required when no template |
| `fields` | [FieldSpec](/concepts/fieldspec) | no | declared selectors |
| `schema` | `json` | no | for the model-assisted reader, every value with its evidence span or refused |
| `source_object` | [SourceObject](/concepts/sourceobject) | yes | declared inline for one scrape, or a registered row |
| `mapping` | [Mapping](/concepts/mapping) | no |  |
| `filter` | [ResourceFilter](/concepts/resourcefilter) | yes |  |
| `as_identity` | `id` | no |  |
| `proxy_policy` | [ProxyPolicy](/concepts/proxypolicy) | no |  |
| `credit_cells` | `[id]` | no |  |
| `asked_by` | `text` | no | who asked for the run, as the caller's kind, or the crawler for an address a walk found; the beat takes what a person or the agent asked before the crawler's own discoveries, oldest first within each |