chore(automation): tighten run.py + envelope.py docstrings

Re-apply the trim style after the prior refactor commit re-introduced
a multi-line docstring on AutomationRun.

- AutomationRun: drop the four-line docstring explaining where
  per-step session ids live; move the note to a single-line inline
  comment right above ``step_results`` where it's actionable.
- AutomationDefinition: drop the design-plan cross-reference; the
  module docstring already establishes what the file is.

No behaviour change.

surfsense_backend/app/automations/persistence/models/run.py

@@ -1,9 +1,4 @@
-"""``automation_runs`` table — immutable per-fire execution record.
-
-Per-step metadata (incl. any LangGraph session id for an ``agent_task`` step)
-lives inside ``step_results[i]``, since a single run may contain zero, one,
-or N agent steps.
-"""
+"""``automation_runs`` table — immutable per-fire execution record."""
 
 from __future__ import annotations
 
@@ -51,6 +46,8 @@ class AutomationRun(BaseModel, TimestampMixin):
 
     trigger_payload = Column(JSONB, nullable=True)
     resolved_inputs = Column(JSONB, nullable=False, server_default="{}")
+    # one entry per executed step; agent_task entries carry their own
+    # `agent_session_id` (LangGraph thread reference) inside this JSONB
     step_results = Column(JSONB, nullable=False, server_default="[]")
     output = Column(JSONB, nullable=True)
     artifacts = Column(JSONB, nullable=False, server_default="[]")