generated: '2026-07-19'
method: searched
source: https://developer.gloat.com/docs/using-gloat-apis
docs:
- https://developer.gloat.com/docs/using-gloat-apis
- https://developer.gloat.com/docs/authentication
conventions:
  base_url:
    pattern: https://{company_slug}.gloat.com/api/{service}/{endpoint}
    per_tenant: true
    staging: https://{company_slug}.gloat-staging.gloat.com/
    note: Every request is scoped to a company slug (tenant) in the hostname.
  authentication:
    customer_apis: Bearer JWT (24h) obtained via client-credentials token endpoint
    talent_marketplace: X-Gloat-API-Key header + X-Tenant-ID header
    ref: authentication/gloat-authentication.yml
  versioning:
    style: uri-path
    versions: [v1, v2]
    example_paths:
    - /api/v2/{company}/...
    - /api/company/v1/...
    - /api/authorization/v1/...
    - /api/candidacy/v1/...
    ref: lifecycle/gloat-lifecycle.yml
  write_semantics:
    upsert_by_natural_key: true
    note: >-
      Many write operations are explicit upserts keyed by the caller-supplied external
      id (user-upsert, job-upsert, candidacy-upsert, project bulk create-or-update,
      learning-items upsert). Re-sending the same external id updates rather than
      duplicates, giving idempotent-by-key sync behavior. Gloat does not document a
      dedicated Idempotency-Key request header.
  bulk_and_multistatus:
    supported: true
    note: >-
      Bulk mutation endpoints (bulk skill-proficiency patch, task-automation-status
      patch) return HTTP 207 Multi-Status with per-item results and per-item status
      codes (200 success, 206 partial, 404 not_found).
  async_jobs:
    pattern: trigger-then-poll
    note: >-
      Resource-intensive file generation/download is asynchronous: a trigger endpoint
      returns a task id, then the download URL is polled by task id (Trigger File
      Download -> Get File Download Url). Data-direct exports return short-lived
      pre-signed S3 URLs.
  quotas_and_credits:
    supported: true
    note: >-
      Certain intelligence endpoints are metered in credits (e.g. Search-by-keyword
      and Personalized opportunity recommendations cost 5 credits per call).
  standards:
    xapi: Learning Action endpoint accepts xAPI-spec statements (batch up to 100).
  transport:
    https_only: true