fix(file-tools): sanitize host/relative cwd override before it reaches container sandbox (#54447) (#54616)

(cherry picked from commit 82132f7911)

Co-authored-by: Tranquil-Flow <66773372+Tranquil-Flow@users.noreply.github.com>

tests/tools/test_container_cwd_sanitize.py

@@ -144,3 +144,114 @@ class TestOverrideCwdSanitizedAtCallSite:
         # RL/benchmark envs set an in-container path; it must pass through.
         cwd = self._run_and_capture_cwd(monkeypatch, "/workspace/task42")
         assert cwd == "/workspace/task42"
+
+
+class TestFileOpsCwdSanitizedAtCallSite:
+    """E2E pin: file tools (_get_file_ops) must sanitize a host/relative cwd
+    override before it reaches _create_environment on a container backend —
+    the same guard the terminal tool got in #50636.  Without it, a Desktop/TUI
+    host cwd (e.g. ``/Users/me/workspace``) leaks straight into
+    ``docker run -w`` and ``search_files`` returns an empty workspace (#54447).
+    """
+
+    def _run_and_capture_cwd(self, monkeypatch, override_cwd, env_type="docker",
+                             config_cwd="/workspace"):
+        """Drive ``_get_file_ops()`` on a container backend with a host-path cwd
+        override registered, and return the cwd that reached
+        ``_create_environment`` (i.e. the cwd passed to ``docker run -w``).
+        """
+        import tools.terminal_tool as tt
+        import tools.file_tools as ft
+
+        captured = {}
+
+        config = {
+            "env_type": env_type,
+            "docker_image": "pytorch/pytorch:latest",
+            "singularity_image": "docker://pytorch/pytorch:latest",
+            "modal_image": "pytorch/pytorch:latest",
+            "daytona_image": "pytorch/pytorch:latest",
+            "cwd": config_cwd,
+            "host_cwd": None,
+            "timeout": 180,
+            "lifetime_seconds": 300,
+            "container_cpu": 1,
+            "container_memory": 5120,
+            "container_disk": 51200,
+            "container_persistent": True,
+            "docker_volumes": [],
+            "docker_env": {},
+            "docker_extra_args": [],
+            "docker_mount_cwd_to_workspace": False,
+            "docker_run_as_host_user": False,
+            "docker_forward_env": [],
+            "modal_mode": "auto",
+            "ssh_host": "",
+            "ssh_user": "",
+            "ssh_port": 22,
+            "ssh_key": "",
+            "ssh_persistent": False,
+            "local_persistent": False,
+        }
+
+        class _DummyEnv:
+            cwd = config_cwd
+
+            def execute(self, *a, **k):
+                return {"output": "", "exit_code": 0}
+
+        def fake_create_environment(env_type, image, cwd, timeout, **kwargs):
+            captured["cwd"] = cwd
+            return _DummyEnv()
+
+        monkeypatch.setattr(tt, "_get_env_config", lambda: config)
+        monkeypatch.setattr(tt, "_start_cleanup_thread", lambda: None)
+        monkeypatch.setattr(tt, "_create_environment", fake_create_environment)
+        # Force a fresh environment build.
+        monkeypatch.setattr(tt, "_active_environments", {})
+        monkeypatch.setattr(tt, "_last_activity", {})
+        monkeypatch.setattr(ft, "_file_ops_cache", {})
+        monkeypatch.setattr(ft, "_last_known_cwd", {})
+
+        task_id = "sess-fileops-host-cwd"
+        tt.register_task_env_overrides(task_id, {"cwd": override_cwd})
+        try:
+            ft._get_file_ops(task_id)
+        finally:
+            tt.clear_task_env_overrides(task_id)
+        return captured.get("cwd")
+
+    def test_macos_host_override_does_not_reach_container(self, monkeypatch):
+        # Desktop/TUI registers /Users/<me>/workspace as the session cwd.
+        cwd = self._run_and_capture_cwd(monkeypatch, "/Users/me/workspace")
+        assert cwd == "/workspace", (
+            f"Host-path cwd override leaked to the container builder: {cwd!r}. "
+            "It must be sanitized back to config['cwd']."
+        )
+
+    def test_posix_home_host_override_does_not_reach_container(self, monkeypatch):
+        cwd = self._run_and_capture_cwd(monkeypatch, "/home/someuser/project")
+        assert cwd == "/workspace"
+
+    def test_windows_host_override_does_not_reach_container(self, monkeypatch):
+        cwd = self._run_and_capture_cwd(monkeypatch, r"C:\Users\someuser")
+        assert cwd == "/workspace"
+
+    def test_relative_cwd_override_does_not_reach_container(self, monkeypatch):
+        cwd = self._run_and_capture_cwd(monkeypatch, "src/app")
+        assert cwd == "/workspace"
+
+    def test_valid_container_override_is_preserved(self, monkeypatch):
+        # RL/benchmark envs set an in-container path; it must pass through.
+        cwd = self._run_and_capture_cwd(monkeypatch, "/workspace/task42")
+        assert cwd == "/workspace/task42"
+
+    def test_host_override_sanitized_on_singularity(self, monkeypatch):
+        cwd = self._run_and_capture_cwd(
+            monkeypatch, "/Users/me/workspace", env_type="singularity")
+        assert cwd == "/workspace"
+
+    def test_host_override_sanitized_on_modal(self, monkeypatch):
+        cwd = self._run_and_capture_cwd(
+            monkeypatch, "/Users/me/workspace", env_type="modal")
+        assert cwd == "/workspace"

tools/file_tools.py

@@ -832,6 +832,8 @@ def _get_file_ops(task_id: str = "default") -> ShellFileOperations:
         _creation_locks,
         _creation_locks_lock,
         _resolve_container_task_id,
+        _is_unusable_container_cwd,
+        _CONTAINER_BACKENDS,
     )
     import time
 
@@ -893,6 +895,26 @@ def _get_file_ops(task_id: str = "default") -> ShellFileOperations:
                 image = ""
 
             cwd = overrides.get("cwd") or _last_known_cwd.get(task_id) or config["cwd"]
+            # Re-apply the container cwd guard that _get_env_config() already
+            # ran on config["cwd"] (see #50636).  A per-task cwd override
+            # registered by the gateway/TUI/ACP for workspace tracking is a
+            # raw host path (e.g. a Desktop session's /Users/<me>/workspace or
+            # C:\\Users\\<me>). On a container backend that reaches
+            # ``docker run -w <host-path>`` and the container starts in a
+            # directory that doesn't exist inside the sandbox, so search_files
+            # and friends silently return empty results (#54447).  Sanitize it
+            # back to the already-validated config["cwd"] so the override can't
+            # bypass the guard.  Valid in-container override paths (RL/benchmark
+            # sandboxes that set cwd to /workspace, /root, etc.) are absolute
+            # non-host paths and pass through untouched.
+            if env_type in _CONTAINER_BACKENDS and _is_unusable_container_cwd(cwd):
+                if cwd != config["cwd"]:
+                    logger.info(
+                        "Ignoring host/relative cwd override %r for %s backend "
+                        "(won't exist in sandbox). Using %r instead.",
+                        cwd, env_type, config["cwd"],
+                    )
+                cwd = config["cwd"]
             logger.info("Creating new %s environment for task %s...", env_type, task_id[:8])
 
             container_config = None