Audit log

The audit log is the ledger of workspace-changing actions — key mints and revocations, product and variant changes, deliveries, webhook endpoints, shipping settings, partner provisioning. It's the same data the portal's activity view renders. The Python SDK wraps one endpoint behind fulkruma.audit_log. For HTTP shapes, see API → Audit log.

Namespace

fulkruma.audit_log     # AuditLogResources

One method. Audit entries are written by the system only — you can't create or delete them via the SDK by design.

Methods

list

fulkruma.audit_log.list(
    *,
    action: str | None = None,
    target_type: str | None = None,
    limit: int | None = None,
    on_behalf_of: str | None = None,
) -> dict

Returns {"entries": [...]}, newest first. Filters:

• action — a prefix of the action name: "api_key." matches api_key.created and api_key.revoked; "product" matches every product.* action.
• target_type — the exact resource type, e.g. "Product", "ApiKey".
• limit — default 100, max 500.

There is no cursor and no date filter: a call returns the newest limit entries that match. To reach further back, narrow action / target_type.

result = fulkruma.audit_log.list(action="api_key.", limit=50)
for e in result["entries"]:
    print(e["createdAt"], e["actorType"], e["actorId"], e["action"], e["targetId"])

Action names in use:

• api_key.created / api_key.revoked
• product.created / product.updated / product.archived
• variant.created / variant.updated / variant.archived
• delivery.created / delivery.extend / delivery.reset-downloads / delivery.revoke
• webhook.created / webhook.updated / webhook.deleted
• shipping.origin_updated / shipping.config_updated
• partner.workspace_provisioned
• storlaunch.product.synced

What triggers each is in API → Audit log.

Types

Each entry is a plain dict:

{
    "id": "clx...",
    "accountId": "acc_...",
    "actorType": "user" | "api_key" | "system",
    "actorId": "..." | None,       # Huudis user ID, or the workspace ID for an API-key call
    "actorEmail": "..." | None,
    "action": "api_key.created",
    "targetType": "ApiKey" | None,
    "targetId": "..." | None,
    "ip": "..." | None,
    "userAgent": "..." | None,
    "before": {...} | None,
    "after": {...} | None,
    "metadata": {...},
    "createdAt": "2026-05-12T10:00:00.000Z",
}

before / after carry the fields the action changed and vary by action — treat them as unknown-typed and pick fields out at the per-action call site. See API → Audit log.

Common patterns

Tail the most recent activity. For a "what just happened?" pane:

recent = fulkruma.audit_log.list(limit=25)["entries"]   # already newest-first

Export what the log holds for one resource type. Take the most the endpoint returns in one call:

import csv

entries = fulkruma.audit_log.list(target_type="Product", limit=500)["entries"]
with open("product-audit.csv", "w", newline="") as f:
    w = csv.writer(f)
    w.writerow(["createdAt", "actorId", "action", "targetId"])
    for e in entries:
        w.writerow([e["createdAt"], e["actorId"], e["action"], e["targetId"]])

If a resource type has more than 500 entries, only the newest 500 come back; split the export by action prefix to reach more.

Several actions. action is a single prefix per call — loop client-side if you need several:

def entries_for_actions(fulkruma, prefixes: list[str]):
    out = []
    for p in prefixes:
        out.extend(fulkruma.audit_log.list(action=p, limit=500)["entries"])
    return sorted(out, key=lambda e: e["createdAt"], reverse=True)

key_and_webhook = entries_for_actions(fulkruma, ["api_key.", "webhook."])

Errors

err.status	err.code	Cause
403	NO_ACCOUNT	The credentials resolve to no workspace.

Unknown action / target_type values aren't errors — they match nothing and return an empty list.

Next

• API keys — the most-audited resource.
• Webhooks — real-time events for shipments, stock and licenses (the audit log itself emits none).
• API → Audit log — HTTP reference, including the full action table.