generated: '2026-07-19'
method: searched
source: >-
  https://documentation.lakesidesoftware.com/docs/systrack-api-1,
  https://documentation.lakesidesoftware.com/docs/ingest-api-user-guide,
  https://documentation.lakesidesoftware.com/docs/signed-secret-authentication-for-webhooks,
  https://documentation.lakesidesoftware.com/docs/systrack-versions-and-life-cycle
description: >-
  Cross-cutting runtime semantics for the SysTrack API surface, harvested from Lakeside's public
  documentation. Coverage is partial by necessity: the full API reference sits behind customer
  sign-in, so conventions that are only described there (header names, pagination parameters, the
  error envelope) are recorded as unknown rather than guessed.
base_url: https://cloud.lakesidesoftware.com/api/
api_style: REST over HTTPS, JSON request and response bodies
deployment_models: [cloud (SaaS tenant), on-premises]
authentication:
  scheme: SysTrack API key (Read, Read/Write, Ingest, SysTrack AI Partner variants)
  user_signin: Azure AD B2C / Microsoft Entra ID, or SAML SSO
  mcp: OAuth 2.0 with Microsoft Entra ID
  detail: authentication/lakeside-software-authentication.yml
idempotency:
  supported: false
  mechanism: null
  note: >-
    No idempotency-key header or replay-safety mechanism is documented on any public SysTrack page.
    The Ingest API instead achieves safety through an explicit staged workflow (upload, diff review,
    then a separate process call) rather than through client-supplied idempotency keys. Recorded as
    unsupported rather than asserted — no Idempotency pointer is wired in apis.yml.
staged_write_workflow:
  applies_to: Ingest API
  description: >-
    The Ingest API deliberately splits a write into discrete, reviewable steps so that schema or data
    drift is surfaced before anything is committed.
  steps:
  - Define the target schema for users or devices (JSON) and send it to SysTrack.
  - Send a CSV file matching the defined schema.
  - SysTrack analyzes the upload and reports differences between the CSV and the current schema.
  - Review the differences and decide how to proceed.
  - Explicitly call the process endpoint to start ingestion.
  - SysTrack ingests the data and exposes it via RPT_EnrichedDevices and RPT_EnrichedUsers.
  - Ingested devices are associated with their WGUID, optimized to run during off-hours.
  docs: https://documentation.lakesidesoftware.com/docs/ingest-api-user-guide
pagination:
  style: unknown
  note: Pagination parameters are documented only in the customer-gated API reference.
field_expansion:
  supported: unknown
metadata:
  supported: true
  mechanism: >-
    Ingest API — up to 2,000 bytes of metadata per CSV row, typed via the schema
    (int, double, datetime, guid, boolean, string).
request_id_tracing:
  header: null
  observed:
  - name: LSHttpAuditSession
    kind: cookie
    note: >-
      Observed on responses from https://cloud.lakesidesoftware.com/api/ — a session/audit
      correlation cookie set by the API gateway. Not a documented client-facing trace header.
  - name: request-context
    kind: response header
    note: Azure Application Insights correlation header observed on the API gateway response.
versioning:
  scheme: product/agent version (e.g. 11.8); no per-request API version header is documented
  detail: lifecycle/lakeside-software-lifecycle.yml
error_envelope:
  format: unknown
  rfc9457: false
  note: >-
    No public error reference or problem-type registry was found; the error catalog lives behind
    customer sign-in. No errors/ artifact was produced rather than inventing problem types.
rate_limiting:
  headers: unknown
  policy: >-
    A "Fair Use Policy" section exists in the SysTrack API course, so limits are asserted, but no
    public numbers or response headers are published.
  detail: authentication/lakeside-software-authentication.yml
webhooks:
  supported: true
  signing: HMAC-SHA256, Base64, x-lakeside-hmac-sha256 header
  detail: asyncapi/lakeside-software-webhooks.yml
tooling:
  documented_clients: [browser, Postman, cURL, PowerShell]
  note: >-
    The SysTrack API course walks through calling the API from a browser, from Postman (including
    "Visualizer endpoints"), and from PowerShell for both cloud and on-premises. These are usage
    patterns, not first-party SDKs — no published client libraries were found.
  detail: packages/lakeside-software-packages.yml
data_constraints:
  ingest:
  - Only CSV files are supported.
  - Up to 2,000 bytes of metadata per row.
  - A CSV file can contain rows for up to 125% of the current device or user count.
  - The schema must be in JSON format.
  - Column names may contain only letters, digits and underscores.
  - Column names must not match SQL reserved words.
  - Supported data types are int, double, datetime, guid, boolean and string.
  - String columns must have a length value greater than 0.
  - Device key column type must be SystemName or Serial.
  - Users are mapped by email as the unique identifier.