Fix validate_tool hang: pass stdin=DEVNULL to the validation subprocess (#218)

* Fix validate_tool hang: pass stdin=DEVNULL to the validation subprocess

`validate_tool` runs each tool's smoke test (e.g. `pi --print "say hi in 5
words or less"`) via `subprocess.run(..., timeout=60)` without setting
`stdin`, so the child inherits ucode's own stdin. In print/headless mode some
agent CLIs (pi among them) still open stdin and block waiting for input/EOF.

When ucode is launched from a non-interactive parent whose stdin is an open
pipe with no EOF — e.g. an agent harness spawning `ucode <tool>` as a
subprocess — the validation command never sees EOF, blocks for the full 60s,
and `validate_tool` returns `(False, "timed out")`. The caller
(`_auto_configure_tool` / configure) then reverts the config and aborts with
"<tool> validation failed — config reverted", even though the tool is
correctly configured and the gateway is reachable.

The validation smoke test is one-shot and never needs interactive input, so
pass `stdin=subprocess.DEVNULL`. This is benign for every tool (an interactive
prompt in a validation run would itself be a bug) and makes validation robust
regardless of how ucode was launched.

Verified on a managed sandbox: `pi --print "say hi"` returns instantly with
`stdin=DEVNULL` but hangs to the timeout when it inherits an open stdin pipe.

Signed-off-by: Edwin He <edwin.he@databricks.com>

* Remove explanatory comment on the validate_tool stdin=DEVNULL kwarg

Per review: the kwarg is self-explanatory; drop the comment block.

---------

Signed-off-by: Edwin He <edwin.he@databricks.com>

Author: Edwinhe03

src/ucode/agents/__init__.py

@@ -466,7 +466,13 @@ def validate_tool(tool: str) -> tuple[bool, str]:
             env = None
     try:
         result = subprocess.run(
-            cmd, check=False, capture_output=True, text=True, timeout=60, env=env
+            cmd,
+            check=False,
+            capture_output=True,
+            text=True,
+            timeout=60,
+            env=env,
+            stdin=subprocess.DEVNULL,
         )
         if result.returncode == 0:
             return True, ""

tests/test_agents_init.py

@@ -567,3 +567,37 @@ def test_low_verbosity_omits_panels(self, monkeypatch, capsys):
         assert "Ready" not in out
         # Per-tool success line is still printed.
         assert "Codex is working" in out
+
+
+class TestValidateTool:
+    def test_runs_validate_command_with_stdin_devnull(self, monkeypatch):
+        # Regression guard: the validation smoke test must never inherit the
+        # caller's stdin, or it hangs to the timeout when ucode is launched
+        # from a non-interactive parent whose stdin is an open pipe.
+        captured: dict = {}
+
+        def fake_run(cmd, **kwargs):
+            captured["cmd"] = cmd
+            captured["kwargs"] = kwargs
+            return subprocess.CompletedProcess(cmd, 0, stdout="", stderr="")
+
+        monkeypatch.setattr("ucode.agents.subprocess.run", fake_run)
+        monkeypatch.setattr(agents_mod, "load_state", lambda: {})
+
+        ok, err = agents_mod.validate_tool("codex")
+
+        assert ok is True
+        assert err == ""
+        assert captured["kwargs"].get("stdin") is subprocess.DEVNULL
+
+    def test_reports_timed_out_on_timeout(self, monkeypatch):
+        def fake_run(cmd, **kwargs):
+            raise subprocess.TimeoutExpired(cmd, 60)
+
+        monkeypatch.setattr("ucode.agents.subprocess.run", fake_run)
+        monkeypatch.setattr(agents_mod, "load_state", lambda: {})
+
+        ok, err = agents_mod.validate_tool("codex")
+
+        assert ok is False
+        assert err == "timed out"