refactor: rename the application package to docsgpt

The backend import package is now docsgpt, the name it will carry on PyPI;
application was far too generic to install into anyone's site-packages.
git mv plus a mechanical rewrite of every import, dotted string and path
reference: 734 Python files, the compose files, Dockerfile, workflows, docs,
setup scripts, devcontainer, k8s manifests, vscode config, pytest and coverage
config, .gitignore. Behaviour is unchanged.

Kept for one release:
- A top-level application package whose meta-path finder resolves
  application.x.y to the already-imported docsgpt.x.y object, so old imports
  and entry points (celery -A application.app.celery,
  uvicorn application.asgi:asgi_app) keep working with a FutureWarning.
- Celery registers every application.* task name as an alias of its
  docsgpt.* task on start-up, so messages queued by the previous release still
  run. The redbeat key prefix moves to redbeat:docsgpt:v2: so schedule entries
  the previous release wrote are left unread instead of firing twice.

The backend image builds from the repository root (docker build -f
docsgpt/Dockerfile .) so it can ship the alias package; a root .dockerignore
allow-lists docsgpt/ and application/ and keeps caches, local data, .env
files, the sample index files and the Dockerfile out. Compose and the image
workflows point at the new context.

docsgpt/agents/workflows/node_agent.py

@@ -0,0 +1,81 @@
+"""Workflow Node Agents - defines specialized agents for workflow nodes."""
+
+from typing import Dict, List, Optional, Type
+
+from docsgpt.agents.agentic_agent import AgenticAgent
+from docsgpt.agents.base import BaseAgent
+from docsgpt.agents.classic_agent import ClassicAgent
+from docsgpt.agents.research_agent import ResearchAgent
+from docsgpt.agents.workflows.schemas import AgentType
+
+
+class _WorkflowNodeMixin:
+    """Common __init__ for all workflow node agents."""
+
+    def __init__(
+        self,
+        endpoint: str,
+        llm_name: str,
+        model_id: str,
+        api_key: str,
+        tool_ids: Optional[List[str]] = None,
+        **kwargs,
+    ):
+        super().__init__(
+            endpoint=endpoint,
+            llm_name=llm_name,
+            model_id=model_id,
+            api_key=api_key,
+            **kwargs,
+        )
+        # Scope the executor to exactly the node's configured tools. Agents
+        # fetch their toolset via ``tool_executor.get_tools()``, so the scope
+        # must live on the executor — it resolves builtin synthetic ids
+        # (Artifact / Code Executor / Read Document) and ``user_tools`` rows
+        # alike, and an empty list means the node's LLM gets no tools.
+        self.tool_executor.allowed_tool_ids = [str(t) for t in (tool_ids or [])]
+
+
+class WorkflowNodeClassicAgent(_WorkflowNodeMixin, ClassicAgent):
+    pass
+
+
+class WorkflowNodeAgenticAgent(_WorkflowNodeMixin, AgenticAgent):
+    pass
+
+
+class WorkflowNodeResearchAgent(_WorkflowNodeMixin, ResearchAgent):
+    pass
+
+
+class WorkflowNodeAgentFactory:
+
+    _agents: Dict[AgentType, Type[BaseAgent]] = {
+        AgentType.CLASSIC: WorkflowNodeClassicAgent,
+        AgentType.REACT: WorkflowNodeClassicAgent,  # backwards compat
+        AgentType.AGENTIC: WorkflowNodeAgenticAgent,
+        AgentType.RESEARCH: WorkflowNodeResearchAgent,
+    }
+
+    @classmethod
+    def create(
+        cls,
+        agent_type: AgentType,
+        endpoint: str,
+        llm_name: str,
+        model_id: str,
+        api_key: str,
+        tool_ids: Optional[List[str]] = None,
+        **kwargs,
+    ) -> BaseAgent:
+        agent_class = cls._agents.get(agent_type)
+        if not agent_class:
+            raise ValueError(f"Unsupported agent type: {agent_type}")
+        return agent_class(
+            endpoint=endpoint,
+            llm_name=llm_name,
+            model_id=model_id,
+            api_key=api_key,
+            tool_ids=tool_ids,
+            **kwargs,
+        )