fix(plugin): persist v4 runtime state and tolerate runner registration gaps (#2774)

Persist the v4 plugin runtime state so recreating the runtime container no
longer discards installed plugins' artifacts, dependency environments and
installation-private data, and stop a transient Runner registry miss during
steering from dropping the message.

- docker-compose: bind-mount ./data/plugin-runtime for the plugin runtime
  alongside ./data/plugins.
- kubernetes: add the langbot-plugin-runtime-state claim (5Gi) mounted at
  /app/data/plugin-runtime, kept separate from the plugin data claim so
  runtime state does not land inside the plugin directory.
- chat handler: treat RunnerNotFoundError while claiming steering as
  "steering unavailable" and continue with a normal run; every other error
  keeps propagating. The runner error imports move to module scope, which
  the new early reference requires.
- tests: cover the steering fallback, the claimed-steering interrupt, and
  the lasting-gap path that still reports runner-not-found.

The steering fallback buys one extra uncached registry lookup while a
runtime reconnects; it does not recover a gap that outlasts the message.

docker/docker-compose.yaml

@@ -9,6 +9,11 @@ services:
     container_name: langbot_plugin_runtime
     volumes:
       - ./data/plugins:/app/data/plugins
+      # Persist the v4 plugin artifact cache, dependency environments, and
+      # per-installation data. Without this, recreating the container throws
+      # away every installed plugin's environment and private data while
+      # data/plugins survives.
+      - ./data/plugin-runtime:/app/data/plugin-runtime
     ports:
       - 5401:5401
     restart: on-failure

docker/kubernetes.yaml

@@ -81,6 +81,25 @@ spec:
   # Uncomment and modify if you have a specific StorageClass
   # storageClassName: your-storage-class
 
+---
+# PersistentVolumeClaim for the v4 Plugin Runtime state
+# Holds the plugin artifact cache, per-plugin dependency environments, and
+# per-installation data. It must survive Pod restarts: rebuilding it forces
+# every installed plugin to be re-downloaded and its environment rebuilt.
+apiVersion: v1
+kind: PersistentVolumeClaim
+metadata:
+  name: langbot-plugin-runtime-state
+  namespace: langbot
+spec:
+  accessModes:
+    - ReadWriteOnce
+  resources:
+    requests:
+      storage: 5Gi
+  # Uncomment and modify if you have a specific StorageClass
+  # storageClassName: your-storage-class
+
 ---
 # ConfigMap for environment configuration
 apiVersion: v1
@@ -160,6 +179,8 @@ spec:
         volumeMounts:
         - name: plugin-data
           mountPath: /app/data/plugins
+        - name: plugin-runtime-state
+          mountPath: /app/data/plugin-runtime
         resources:
           requests:
             memory: "512Mi"
@@ -189,6 +210,9 @@ spec:
       - name: plugin-data
         persistentVolumeClaim:
           claimName: langbot-plugin-runtime-data
+      - name: plugin-runtime-state
+        persistentVolumeClaim:
+          claimName: langbot-plugin-runtime-state
       restartPolicy: Always
 
 ---