robustness: vision/depth render off the event loop (deepcopy snapshot), chat auto-reconnect + friendly errors, demo catches connection failures

.playwright-mcp/console-2026-08-08T09-11-01-487Z.log

@@ -95,3 +95,4 @@
 [  183147ms] [ERROR] WebSocket connection to 'ws://127.0.0.1:8001/' failed: Error in connection establishment: net::ERR_CONNECTION_REFUSED @ http://127.0.0.1:8000/:40
 [  188179ms] [ERROR] WebSocket connection to 'ws://127.0.0.1:8001/' failed: Error in connection establishment: net::ERR_CONNECTION_REFUSED @ http://127.0.0.1:8000/:40
 [  192874ms] [ERROR] WebSocket connection to 'ws://127.0.0.1:8001/' failed: Error in connection establishment: net::ERR_CONNECTION_REFUSED @ http://127.0.0.1:8000/:40
+[ 1348483ms] [ERROR] WebSocket connection to 'ws://127.0.0.1:8001/' failed: Error in connection establishment: net::ERR_CONNECTION_REFUSED @ http://127.0.0.1:8000/:40

testbed/bridge.py

@@ -11,6 +11,7 @@ from __future__ import annotations
 
 import asyncio
 import base64
+import copy
 import io
 import math
 from typing import Any
@@ -206,11 +207,18 @@ class RoomBridge(Bridge):
         )
         async def vision() -> VisionOutput:
             tick = self._tick()
-            img, _ = renderer.render(world)
-            buf = io.BytesIO()
-            img.save(buf, format="PNG")
+            # Render off the event loop so frames never stall the bridge under
+            # load; deepcopy keeps the snapshot consistent with this tool call.
+            snapshot = copy.deepcopy(world)
+            img, _ = await asyncio.to_thread(renderer.render, snapshot)
+
+            def _encode() -> str:
+                buf = io.BytesIO()
+                img.save(buf, format="PNG")
+                return base64.b64encode(buf.getvalue()).decode("ascii")
+
             return VisionOutput(
-                png_b64=base64.b64encode(buf.getvalue()).decode("ascii"),
+                png_b64=await asyncio.to_thread(_encode),
                 width=VISION_WIDTH,
                 height=VISION_HEIGHT,
                 tick=tick,
@@ -222,7 +230,8 @@ class RoomBridge(Bridge):
         )
         async def depth(max_depth: float = 10.0) -> DepthOutput:
             tick = self._tick()
-            _, grid = renderer.render(world, max_depth=max_depth)
+            snapshot = copy.deepcopy(world)
+            _, grid = await asyncio.to_thread(renderer.render, snapshot, max_depth)
             return DepthOutput(
                 width=DEPTH_WIDTH,
                 height=DEPTH_HEIGHT,

testbed/chat.py

@@ -191,12 +191,34 @@ async def chat_loop(client: AICCClient, manifest, args: argparse.Namespace) -> i
 
 
 async def run(args: argparse.Namespace) -> int:
-    transport = WebSocketClientTransport(args.url)
-    async with AICCClient(transport) as client:
-        manifest = await client.handshake()
-        print(f"[handshake] session {manifest.session_id} world {manifest.world.name}")
-        print(f"[handshake] tools: {[t.id for t in manifest.tools]}")
-        return await chat_loop(client, manifest, args)
+    """Run the chat session, reconnecting to the bridge if the connection dies."""
+    attempts = 0
+    while True:
+        try:
+            transport = WebSocketClientTransport(args.url)
+            async with AICCClient(transport) as client:
+                manifest = await client.handshake()
+                print(
+                    f"[handshake] session {manifest.session_id} world {manifest.world.name}"
+                )
+                print(f"[handshake] tools: {[t.id for t in manifest.tools]}")
+                return await chat_loop(client, manifest, args)
+        except asyncio.CancelledError:
+            raise
+        except KeyboardInterrupt:
+            raise
+        except Exception as exc:  # noqa: BLE001 - connection lost: reconnect
+            attempts += 1
+            print(f"\n[chat] connection lost ({type(exc).__name__}: {exc})")
+            if attempts >= 3:
+                print(
+                    "[chat] giving up after 3 attempts — is the bridge running? (scripts/run_bridge.sh)"
+                )
+                return 1
+            print(
+                "[chat] reconnecting in 2 seconds… (the room keeps its state on the bridge)"
+            )
+            await asyncio.sleep(2.0)
 
 
 def main() -> int:

testbed/demo.py

@@ -152,7 +152,9 @@ async def run_llm_agent(
         log=lambda role, msg: print(f"[{role}] {msg}"),
     )
     try:
-        return await run_llm_agent_loop(controller, max_steps, log=print_log, recorder=recorder)
+        return await run_llm_agent_loop(
+            controller, max_steps, log=print_log, recorder=recorder
+        )
     except RuntimeError as exc:
         return {"steps": 0, "tool_calls": 0, "result": str(exc), "interacted": False}
 
@@ -398,7 +400,14 @@ def main() -> int:
         "into this directory, plus demo.gif animation and demo_summary.png",
     )
     args = parser.parse_args()
-    summary = asyncio.run(run_demo(args))
+    try:
+        summary = asyncio.run(run_demo(args))
+    except KeyboardInterrupt:
+        return 2
+    except Exception as exc:  # noqa: BLE001 - friendly failure instead of a traceback
+        print(f"\n[demo] failed: {type(exc).__name__}: {exc}")
+        print("[demo] is the bridge running?  scripts/run_bridge.sh")
+        return 1
     return 0 if summary.get("interacted") else 1