EU Deforestation Regulation (EUDR)

There are two surfaces, and the difference is just lifecycle state:

• Audit Trail (/audit-trail/*) — the working draft. You create a record from a set of plots, refine it (exclude plots, add justifications, set the operator and commodity blocks, manage suppliers and report fields), cut immutable versions, and export a preview report. Nothing here is filed.
• DDS (/dds/*) — the filed artifact. When you submit an audit record's current version it is sent to TRACES and becomes a DDS, appearing in the submitted lists.

build & refine                    file                      read back
──────────────                    ────                      ─────────
POST /audit-trail  ──refine──▶  POST /dds/:id/submit  ──▶  GET /dds
  (or fast path)                                           GET /dds/:id

Path	When to use it	How
Fast path	"I just uploaded plots/facilities and want the audit record built automatically." Best for straight-through, one-commodity uploads.	Set create_audit_trail: true (with metrics: ["deforestation"]) on the ingestion upload. The worker auto-creates the record once processing finishes; poll the ingestion task's audit_trail_task_id.
Explicit path	"I want to choose exactly which plots go in, supply the operator/commodity details up front, or split one upload into several records."	Create Audit Trail — POST /audit-trail with a filter selecting the plots plus optional operator, commodity, and grouping dimensions.

Both paths produce the same kind of record and converge on the same audit-trail and DDS endpoints. The fast path uses the org's default operator and an empty commodity block (set the commodity via Update DDS Record before submit). Record creation is asynchronous in both cases — you receive a dds_record_task; poll Get Audit Trail Task until it completes and carries the new record_id.

For runnable, end-to-end recipes see the EUDR Workflows collection (Fast Path from Plots / Facilities, and the Comprehensive explicit walkthrough).

1. Get plots in with a deforestation run — via ingestion (fast path) or any earlier upload.
2. Create the audit record — fast path automatically, or POST /audit-trail with an explicit plot filter.
3. Refine it — exclude irrelevant plots, attach justifications / compliance overrides, set the operator + commodity blocks, pin suppliers, add report data fields. Every edit is captured in the append-only Audit Log.
4. Preview — export the Report (JSON or GeoJSON) to review the package before filing.
5. Submit — record the operator's Authorisation, then Submit DDS. The record becomes a filed DDS.
6. Reprocess when sourcing changes — Reprocess cuts a new version against fresh assessments / a new end_date.

Folder / endpoint	What it's for
Create Audit Trail	Start a new audit record from a plot filter (the explicit path)
Preview Audit Trail Size	Count the plots a create would resolve, before committing to the run
Tasks	Poll async create/version/submit tasks — one by id (Get Audit Trail Task) or the org-wide feed (List All Audit Trail Tasks)
Audit Trail	The working surface — read, refine, version, and report on a record
Due Diligence Statement	The filed surface — submit a record and read back submitted DDS records

• Reference lookups only: eudr:reference permits GET /regulations/eudr/scope, /scope/entries and /species. It grants no audit-trail or DDS access.
• Scopes: eudr:read for reads (including the reference lookups), eudr:write for create/edit/submit.
• Multi-org: a record is visible only to the org that owns it (dds_record.organization_id ∈ effectiveOrgIds); records from another org return 404.