refactor(agents): move permissions to app/agents/shared (slice 4a)

Relocate the permission evaluator (wildcard matcher + rule evaluation) to the
shared kernel and flip 43 non-frozen importers. A re-export shim remains at
new_chat/permissions.py for the frozen single-agent stack (chat_deepagent and
subagents/{config,providers/linear,providers/slack}); it will be removed when
that stack is retired.

surfsense_backend/app/agents/multi_agent_chat/middleware/shared/permissions/ask/payload.py

@@ -13,7 +13,7 @@ from app.agents.multi_agent_chat.subagents.shared.hitl.wire import (
     SURFSENSE_DECISION_APPROVE_ALWAYS,
     build_lc_hitl_payload,
 )
-from app.agents.new_chat.permissions import Rule
+from app.agents.shared.permissions import Rule
 
 PERMISSION_ASK_INTERRUPT_TYPE = "permission_ask"
 

surfsense_backend/app/agents/multi_agent_chat/middleware/shared/permissions/ask/request.py

@@ -16,7 +16,7 @@ from typing import Any
 from langchain_core.tools import BaseTool
 from langgraph.types import interrupt
 
-from app.agents.new_chat.permissions import Rule
+from app.agents.shared.permissions import Rule
 from app.observability import metrics as ot_metrics, otel as ot
 
 from .decision import normalize_permission_decision

surfsense_backend/app/agents/multi_agent_chat/middleware/shared/permissions/deny.py

@@ -12,7 +12,7 @@ from typing import Any
 from langchain_core.messages import ToolMessage
 
 from app.agents.shared.errors import StreamingError
-from app.agents.new_chat.permissions import Rule
+from app.agents.shared.permissions import Rule
 
 
 def build_deny_message(tool_call: dict[str, Any], rule: Rule) -> ToolMessage:

surfsense_backend/app/agents/multi_agent_chat/middleware/shared/permissions/middleware/core.py

@@ -27,7 +27,7 @@ from langchain_core.tools import BaseTool
 from langgraph.runtime import Runtime
 
 from app.agents.shared.errors import CorrectedError, RejectedError
-from app.agents.new_chat.permissions import Ruleset
+from app.agents.shared.permissions import Ruleset
 from app.services.user_tool_allowlist import TrustedToolSaver
 
 from ..ask.edit import merge_edited_args

surfsense_backend/app/agents/multi_agent_chat/middleware/shared/permissions/middleware/evaluation.py

@@ -16,7 +16,7 @@ from __future__ import annotations
 import logging
 from typing import Any
 
-from app.agents.new_chat.permissions import (
+from app.agents.shared.permissions import (
     Rule,
     RuleAction,
     Ruleset,

surfsense_backend/app/agents/multi_agent_chat/middleware/shared/permissions/middleware/factory.py

@@ -28,7 +28,7 @@ from collections.abc import Sequence
 from langchain_core.tools import BaseTool
 
 from app.agents.shared.feature_flags import AgentFeatureFlags
-from app.agents.new_chat.permissions import Rule, Ruleset
+from app.agents.shared.permissions import Rule, Ruleset
 from app.services.user_tool_allowlist import TrustedToolSaver
 
 from .core import PermissionMiddleware

surfsense_backend/app/agents/multi_agent_chat/middleware/shared/permissions/middleware/ruleset_view.py

@@ -9,7 +9,7 @@ newly-promoted rules apply to subsequent calls.
 
 from __future__ import annotations
 
-from app.agents.new_chat.permissions import Ruleset, aggregate_action, evaluate_many
+from app.agents.shared.permissions import Ruleset, aggregate_action, evaluate_many
 
 
 def all_rulesets(

surfsense_backend/app/agents/multi_agent_chat/middleware/shared/permissions/middleware/runtime_promote.py

@@ -7,7 +7,7 @@ is the streaming layer's job — this module keeps the in-memory copy only.
 
 from __future__ import annotations
 
-from app.agents.new_chat.permissions import Rule, Ruleset
+from app.agents.shared.permissions import Rule, Ruleset
 
 
 def persist_always(