Compass Properties and Details Scraper

This Actor processes Compass property pages only when Compass has expressly authorized automated collection and the intended use in writing. Compass's Terms of Service prohibit crawling, spidering, scraping, and harvesting Content; they also limit ordinary use of Content to personal, non-commercial purposes. The current workspace has no recorded written authorization.

The Actor makes no source requests by default. Set contentUseAuthorized to true only after confirming that written authorization covers this automated use. This is an operator attestation; it does not grant permission. The Actor accepts explicit HTTPS URLs on compass.com or www.compass.com, refuses redirects, applies a 15-second request timeout and a 5 MiB response limit, and processes only the provided URLs. It does not discover or paginate search results.

For an authorized page, the current parser reads supported property facts from Schema.org JSON-LD and emits a row only when it finds a canonical Compass URL and at least one property fact. The dataset fields are recordType, listingId, address, price, beds, baths, propertyType, url, sourceUrl, and scrapedAt; absent values are omitted. A URL or identifier alone is not enough to create a record. Parser tests use synthetic data and do not establish current Compass extraction.

The default policy skip reports status: SUCCEEDED and runStatus: SKIPPED_POLICY, with zero stored items. A valid empty result is SUCCEEDED / COMPLETED; a failure with zero stored records is FAILED / FAILED; a failure after one or more records were stored is SUCCEEDED / PARTIAL. Failure codes and counts stay in the SUMMARY key-value record. The Actor does not bypass access controls, bot challenges, or other restrictions.

Dependency security

The locked dependency graph overrides basic-ftp with the patched 6.2.1 release. The SDK’s unpatched HTTP cache policy is replaced with the readable local vendor/noncaching-http-policy adapter, which refuses cache reads and writes and forces fresh network responses. This is an explicit removal of caching, not a claimed upstream security patch. Source transports, permission gates and request limits are unchanged. npm run test:dependencies checks cross-user cache isolation, imported stale policy rejection, FTP integration methods and adversarial listing parsing. Cloud builds run npm audit --omit=dev --audit-level=low against the installed production graph.