Current service
Job Search API runs collection tasks through REST and the workspace. Results are stored privately in Cloudflare R2 and delivered through authorized access. Public website examples contain illustrative data. Account sign-in, saved request configurations, credential metadata, website traffic, optional analytics and email inquiries are described in the privacy notice. The provider list identifies the services used for those purposes.
Collection boundaries
The API design targets supported web sources with an explicit input and output contract. Source names in the catalog are candidates, not confirmation of live access or endorsement. Capabilities requiring an end user’s personal source-site password or session token are outside the selected product boundary.
Availability, permitted use and technical limits must be reviewed for each source before opening a capability. Describing a source as public does not establish unrestricted rights to every use of its data.
Freshness and caching
The planned contract distinguishes retrieval time from the source’s own update time. source.collectedAt records when TrueFetch collected data; it does not promise the source itself was current. source.cacheHit identifies reuse. Requesting maxAgeS=0 means fetching again rather than using TrueFetch’s cache, not a guarantee of real-time source data.
Cross-tenant reuse is limited to public data permitted for reuse. Tenant results, private parameters and authorization contexts remain isolated. Reading an existing result and creating a new collection are different actions.
Retention and deletion
Results expire at the task’s resultExpiresAt, normally seven days after completion. Short-lived download links do not extend result retention. Billing, security, recovery and backup records can remain after results expire. Result expiry does not delete all account records.
Deletion and suppression must take priority over ordinary expiry, including after restoring backups. A result URL does not promise permanent storage. Deleting a TrueFetch copy cannot recall copies already downloaded by a recipient.
Questions, removal and abuse reports
Email [email protected] with the relevant public URL, the concern and enough information to identify the material. Do not send identity documents or sensitive records in an initial report. We will use the inquiry to assess the request and any necessary follow-up; we do not promise an automated removal service or a fixed response deadline.
For vulnerabilities, use the security reporting guidance.