docs(FI-609): optional team_id on Signals deletion create

specification/resources/signals/models/create_deletion_request.yml

@@ -15,7 +15,11 @@ properties:
   team_id:
     type: integer
     format: int64
-    description: The team whose data should be deleted. Must match the authenticated team.
+    description: >-
+      Numeric team ID. Optional for both `agent` and `inference`. When omitted
+      (or zero), the API uses the team authenticated by your API token
+      (personal access token or OAuth token). When set, it must equal that
+      authenticated team or the request is rejected (`403`).
     example: 12345678
   agent_id:
     type: string
@@ -28,5 +32,4 @@ properties:
     example: 01a10d3f-2c09-764f-99f8-458e521627f3
 required:
   - type
-  - team_id
 additionalProperties: false

specification/resources/signals/signals_create_deletion.yml

@@ -11,13 +11,21 @@ description: |
   - `agent` — deletes Signals conversations for one Harness Runtime environment and
     turns off collection for that environment. Requires `agent_id`, which is the
     environment's Environment Config ID (the environment's `agent_id`). On create, matching
-    conversations are hidden from Signals reads immediately. Wait until the
-    job is `complete` before enabling Signals for the environment again.
+    conversations are hidden from Signals reads immediately. We recommend waiting
+    until the job is `complete` before enabling Signals for the environment again;
+    you can re-enable earlier if needed (new conversations after re-enable are kept).
   - `inference` — deletes Serverless Inference Signals conversations for your
     whole team (team-wide; not limited to one endpoint or model) that were
     collected at or before the deletion request, and turns off team inference
     collection (`source=inference`). Do not send `agent_id`. On create,
-    matching conversations are hidden from Signals reads immediately.
+    matching conversations are hidden from Signals reads immediately. The same
+    re-enable guidance applies to team inference collection.
+
+  `team_id` is optional for both types (including `inference`). When you omit
+  it, the API uses the team from your API token (PAT or OAuth). When you send
+  it, it must match that authenticated team or the request returns `403`.
+  `doctl signals deletion create` also defaults `--team-id` from your token
+  when the flag is omitted.
 
   The job runs asynchronously; poll the get endpoint until `status` is
   `complete` or `failed`. DigitalOcean may retry transient processing errors