From e1b640debb203db5ea7aff6351b57a73f6133129 Mon Sep 17 00:00:00 2001
From: Vamsi-2581 <245859299+Vamsi-2581@users.noreply.github.com>
Date: Tue, 27 Oct 2026 21:32:10 +0530
Subject: [PATCH 1/3] fix: return after backgrounded shell command exits

---
 src/stirrup/tools/code_backends/local.py | 46 +++++++++++-------------
 1 file changed, 20 insertions(+), 26 deletions(-)

diff --git a/src/stirrup/tools/code_backends/local.py b/src/stirrup/tools/code_backends/local.py
index 75fee65..c409904 100644
--- a/src/stirrup/tools/code_backends/local.py
+++ b/src/stirrup/tools/code_backends/local.py
@@ -327,36 +327,30 @@ async def run_command(self, cmd: str, *, timeout: int | None = None) -> CommandR
                 # on timeout. Without it, process.kill() only terminates the
                 # root process, leaving grandchildren reparented to init (same
                 # bug shape as the docker orphan leak that motivated this PR).
-                process = await anyio.open_process(
-                    argv if argv is not None else ["bash", "-c", cmd],
-                    stdout=subprocess.PIPE,
-                    stderr=subprocess.PIPE,
-                    cwd=self._temp_dir,
-                    start_new_session=True,
-                )
-
-                # Read all output from streams concurrently
-                stdout_chunks: list[bytes] = []
-                stderr_chunks: list[bytes] = []
-
-                async def read_stdout() -> None:
-                    if process.stdout:
-                        stdout_chunks.extend([chunk async for chunk in process.stdout])
-
-                async def read_stderr() -> None:
-                    if process.stderr:
-                        stderr_chunks.extend([chunk async for chunk in process.stderr])
-
-                async with anyio.create_task_group() as tg:
-                    tg.start_soon(read_stdout)
-                    tg.start_soon(read_stderr)
+                # Use regular temporary files instead of pipes. A backgrounded
+                # shell command inherits stdout/stderr, so pipe readers cannot see
+                # EOF when the shell exits and would keep this call waiting for the
+                # background process. Files preserve the shell's output while letting
+                # us wait only for the command process itself.
+                with tempfile.TemporaryFile() as stdout_file, tempfile.TemporaryFile() as stderr_file:
+                    process = await anyio.open_process(
+                        argv if argv is not None else ["bash", "-c", cmd],
+                        stdout=stdout_file,
+                        stderr=stderr_file,
+                        cwd=self._temp_dir,
+                        start_new_session=True,
+                    )
+                    await process.wait()
 
-                await process.wait()
+                    stdout_file.seek(0)
+                    stderr_file.seek(0)
+                    stdout = stdout_file.read().decode("utf-8", errors="replace")
+                    stderr = stderr_file.read().decode("utf-8", errors="replace")
 
                 return CommandResult(
                     exit_code=process.returncode or 0,
-                    stdout=b"".join(stdout_chunks).decode("utf-8", errors="replace"),
-                    stderr=b"".join(stderr_chunks).decode("utf-8", errors="replace"),
+                    stdout=stdout,
+                    stderr=stderr,
                 )
 
         except TimeoutError:

From d198a00a85df8cea95113c019bd090b410942982 Mon Sep 17 00:00:00 2001
From: Vamsi-2581 <245859299+Vamsi-2581@users.noreply.github.com>
Date: Tue, 27 Oct 2026 21:32:17 +0530
Subject: [PATCH 2/3] test: cover backgrounded compound shell commands

---
 tests/test_local_execution.py | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/tests/test_local_execution.py b/tests/test_local_execution.py
index 97558fd..688f1cc 100644
--- a/tests/test_local_execution.py
+++ b/tests/test_local_execution.py
@@ -60,6 +60,18 @@ async def test_run_command_timeout(self) -> None:
             assert result.error_kind == "timeout"
             assert "timed out" in result.stderr.lower()
 
+    async def test_run_command_backgrounded_compound_command_returns_without_waiting(self) -> None:
+        """A backgrounded command list must not keep captured output streams open."""
+        provider = LocalCodeExecToolProvider()
+
+        async with provider as _:
+            result = await provider.run_command('echo "hi" && sleep 2 > /dev/null 2>&1 &', timeout=1)
+
+        assert result.exit_code == 0
+        assert result.stdout == "hi\n"
+        assert result.stderr == ""
+        assert result.error_kind is None
+
     async def test_run_command_timeout_kills_descendants(self) -> None:
         """Regression: a timed-out command must not leave orphaned grandchildren.
 

From 4c0413d0e0e4f0b7d3f3ae27f4f56a3ffea51067 Mon Sep 17 00:00:00 2001
From: Vamsi-2581 <245859299+Vamsi-2581@users.noreply.github.com>
Date: Tue, 27 Oct 2026 21:32:55 +0530
Subject: [PATCH 3/3] test: synchronize background command output capture

---
 tests/test_local_execution.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/test_local_execution.py b/tests/test_local_execution.py
index 688f1cc..33e39fa 100644
--- a/tests/test_local_execution.py
+++ b/tests/test_local_execution.py
@@ -65,7 +65,7 @@ async def test_run_command_backgrounded_compound_command_returns_without_waiting
         provider = LocalCodeExecToolProvider()
 
         async with provider as _:
-            result = await provider.run_command('echo "hi" && sleep 2 > /dev/null 2>&1 &', timeout=1)
+            result = await provider.run_command('echo "hi" && touch ready && sleep 2 > /dev/null 2>&1 & while [ ! -f ready ]; do sleep 0.01; done', timeout=1)
 
         assert result.exit_code == 0
         assert result.stdout == "hi\n"