fix: Modal sandbox eval infra (9 fixes for TBLite baseline)

Fixes discovered while running TBLite baseline evaluation:

1. ephemeral_disk param not supported in modal 1.3.5 - check before passing
2. Modal legacy image builder requires working pip - add ensurepip fix via
   setup_dockerfile_commands to handle task images with broken pip
3. Host cwd leaked into Modal sandbox - add /home/ to host prefix check
4. Tilde ~ not expanded by subprocess.run(cwd=) in sandboxes - use /root
5. install_pipx must stay True for swerex-remote to be available

Dependencies also needed (not in this commit):
- git submodule update --init mini-swe-agent
- uv pip install swe-rex boto3

tools/terminal_tool.py

@@ -463,7 +463,7 @@ def _get_env_config() -> Dict[str, Any]:
     if env_type == "local":
         default_cwd = os.getcwd()
     else:
-        default_cwd = "~"
+        default_cwd = "/root"
     
     # Read TERMINAL_CWD but sanity-check it for container backends.
     # If the CWD looks like a host-local path that can't exist inside a
@@ -553,7 +553,12 @@ def _create_environment(env_type: str, image: str, cwd: str, timeout: int,
         if memory > 0:
             sandbox_kwargs["memory"] = memory
         if disk > 0:
-            sandbox_kwargs["ephemeral_disk"] = disk
+            try:
+                import inspect, modal
+                if "ephemeral_disk" in inspect.signature(modal.Sandbox.create).parameters:
+                    sandbox_kwargs["ephemeral_disk"] = disk
+            except Exception:
+                pass
         
         return _ModalEnvironment(
             image=image, cwd=cwd, timeout=timeout,