Files
hermes-agent/tests/tools/test_execute_code_lifecycle_guard_deadline.py
alt-glitch 782e0b42c9 refactor(code-files): delete unreached sandbox gate, checkpoint helpers, py shell linter
Removed. A search for callers by name, attribute and string (Python, TypeScript,
docs, the plugin API list and the frozen old-updater list) finds no supported
path that reaches any of them:

- tools/code_execution_tool.py::SANDBOX_AVAILABLE (constant True) and its
  two `if not SANDBOX_AVAILABLE` branches; the duplicated gateway-lifecycle
  comment in execute_code
- tests: test_windows_returns_error (only flipped the constant) and the
  constant's monkeypatch in test_execute_code_lifecycle_guard_deadline
- tools/checkpoint_manager.py: the no-op `store = _store_path(base)` /
  `_ = store` pair in _migrate_legacy_store; unused imports
  clear_stale_tmp_packs, rmtree_readonly, Dict, List, Set, Tuple;
  _list_project_refs (test-only) and its one assertion
- hermes_cli/cli_auto_maintenance.py::_run_checkpoint_auto_maintenance
  (imported by cli.py, never called), its cli.py import, its orphaned
  `threading` import and its no-op test monkeypatch
- tools/file_operations.py::ShellFileOperations._run_python_snippet
- tools/file_operations_lint.py: the '.py' entry in LINTERS (the in-process
  LINTERS_INPROC['.py'] always answers first)

- test_profile_rename_checkpoints: assert the migrate-identity retry deletes the leftover old checkpoint ref (via _ref_tip), replacing the deleted _list_project_refs check.
- docs(checkpoints-and-rollback): auto-prune runs only on the gateway housekeeping tick; the CLI never runs it.
2026-11-16 15:22:48 +05:30

45 lines
2.1 KiB
Python

"""execute_code's gateway-lifecycle identity probe shares the cell deadline (#111922).
The same ``_is_supervised_gateway_process`` probe that wedged ``terminal_tool`` runs
ahead of every ``execute_code`` cell; a probe that renders no verdict must fail closed
within the deadline instead of holding the tool call (and its cron slot) forever.
"""
from __future__ import annotations
import json
import time
import tools.code_execution_tool as cet
import tools.terminal_tool as terminal_module
def test_wedged_lifecycle_probe_returns_bounded_error_without_running(monkeypatch):
from tools import process_registry
monkeypatch.setattr(cet, "_load_config", lambda: {"timeout": 0.05})
monkeypatch.setattr(terminal_module, "_PRE_EXEC_GUARD_MIN_TIMEOUT_S", 0)
monkeypatch.setattr(process_registry, "_is_supervised_gateway_process", lambda: time.sleep(1))
ran: list[str] = []
monkeypatch.setattr(cet, "_get_env_config", lambda: ran.append("env") or {"env_type": "local"}, raising=False)
start = time.monotonic()
result = json.loads(cet.execute_code("print('hi')"))
elapsed = time.monotonic() - start
assert elapsed < 0.5, f"lifecycle probe wedged execute_code for {elapsed:.2f}s"
assert "did not finish" in result["error"]
assert ran == [], "a probe with no verdict must not fail open into execution"
def test_interpreter_kill_in_execute_code_names_the_owned_process_route(monkeypatch):
"""Sibling surface of the terminal guard (#113667): an image-name kill inside a cell gets the same
proc_* / explicit-PID rejection, not the generic lifecycle text."""
from tools import process_registry
monkeypatch.setattr(process_registry, "_is_supervised_gateway_process", lambda: True)
result = json.loads(cet.execute_code('import subprocess; subprocess.run(["pkill", "-9", "python3"])'))
assert "proc_" in result["error"] and "explicit PID" in result["error"]
generic = json.loads(cet.execute_code('import os; os.system("hermes gateway restart")'))
assert "proc_" not in generic["error"] and "cannot restart or stop the gateway" in generic["error"]