147 lines
4.8 KiB
Python
147 lines
4.8 KiB
Python
from __future__ import annotations
|
|
|
|
import json
|
|
from pathlib import Path
|
|
|
|
import pytest
|
|
|
|
from claudecode_support import (
|
|
ElydoraApiHandler,
|
|
managed_handler,
|
|
prepare_fixture,
|
|
run_handler,
|
|
server_base_url,
|
|
start_api_server,
|
|
)
|
|
|
|
|
|
def official_payload(event: str, **overrides: object) -> dict[str, object]:
|
|
value: dict[str, object] = {
|
|
"session_id": "session-1",
|
|
"prompt_id": "302d811d-0d17-41ad-a359-d2cb618fd42b",
|
|
"transcript_path": "/tmp/session-1.jsonl",
|
|
"cwd": "/tmp/project",
|
|
"permission_mode": "default",
|
|
"effort": {"level": "high"},
|
|
"hook_event_name": event,
|
|
"tool_name": "Bash",
|
|
"tool_input": {"command": "pytest", "description": "Run tests"},
|
|
"tool_use_id": "toolu_01ABC123",
|
|
}
|
|
value.update(overrides)
|
|
return value
|
|
|
|
|
|
def encoded(value: object) -> bytes:
|
|
return json.dumps(value, separators=(",", ":")).encode()
|
|
|
|
|
|
def test_runtimes_preserve_success_and_failure_payloads(
|
|
monkeypatch: pytest.MonkeyPatch, tmp_path: Path
|
|
) -> None:
|
|
server = start_api_server()
|
|
try:
|
|
fixture = prepare_fixture(
|
|
monkeypatch, tmp_path, base_url=server_base_url(server)
|
|
)
|
|
fixture.install()
|
|
settings = fixture.settings()
|
|
guard = managed_handler(settings, "PreToolUse")
|
|
success_audit = managed_handler(settings, "PostToolUse")
|
|
failure_audit = managed_handler(settings, "PostToolUseFailure")
|
|
|
|
pre = official_payload("PreToolUse")
|
|
guard_result = run_handler(guard, encoded(pre), fixture)
|
|
success = official_payload(
|
|
"PostToolUse",
|
|
tool_response={
|
|
"stdout": "tests passed",
|
|
"stderr": "",
|
|
"interrupted": False,
|
|
"isImage": False,
|
|
},
|
|
)
|
|
success_result = run_handler(success_audit, encoded(success), fixture)
|
|
failure = official_payload(
|
|
"PostToolUseFailure",
|
|
error="Command exited with non-zero status code 1",
|
|
is_interrupt=False,
|
|
duration_ms=4187,
|
|
)
|
|
failure_result = run_handler(failure_audit, encoded(failure), fixture)
|
|
finally:
|
|
server.shutdown()
|
|
server.server_close()
|
|
|
|
assert guard_result.returncode == 0
|
|
assert guard_result.stdout == b""
|
|
assert guard_result.stderr == b""
|
|
assert success_result.returncode == 0
|
|
assert failure_result.returncode == 0
|
|
assert [request["method"] for request in ElydoraApiHandler.requests] == [
|
|
"GET", "POST", "POST",
|
|
]
|
|
success_operation = ElydoraApiHandler.requests[1]["json"]
|
|
failure_operation = ElydoraApiHandler.requests[2]["json"]
|
|
assert success_operation["payload"] == success
|
|
assert failure_operation["payload"] == failure
|
|
assert success_operation["subject"] == {"session_id": "session-1"}
|
|
assert success_operation["action"] == {"tool": "Bash"}
|
|
assert ElydoraApiHandler.requests[1]["authorization"] == "Bearer token-1"
|
|
assert failure_operation["prev_chain_hash"] == success_operation["chain_hash"]
|
|
|
|
|
|
@pytest.mark.parametrize("status", ["frozen", "revoked"])
|
|
def test_guard_returns_official_exit_code_two(
|
|
monkeypatch: pytest.MonkeyPatch,
|
|
tmp_path: Path,
|
|
status: str,
|
|
) -> None:
|
|
server = start_api_server(status=status)
|
|
try:
|
|
fixture = prepare_fixture(
|
|
monkeypatch, tmp_path, base_url=server_base_url(server)
|
|
)
|
|
fixture.install()
|
|
result = run_handler(
|
|
managed_handler(fixture.settings(), "PreToolUse"),
|
|
encoded(official_payload("PreToolUse")),
|
|
fixture,
|
|
)
|
|
finally:
|
|
server.shutdown()
|
|
server.server_close()
|
|
|
|
assert result.returncode == 2
|
|
assert result.stdout == b""
|
|
assert status.encode() in result.stderr.lower()
|
|
assert b"Tool execution blocked" in result.stderr
|
|
|
|
|
|
def test_runtime_failures_are_observable_and_fail_open(
|
|
monkeypatch: pytest.MonkeyPatch, tmp_path: Path
|
|
) -> None:
|
|
fixture = prepare_fixture(monkeypatch, tmp_path)
|
|
fixture.install()
|
|
settings = fixture.settings()
|
|
guard = managed_handler(settings, "PreToolUse")
|
|
audit = managed_handler(settings, "PostToolUse")
|
|
|
|
guard_result = run_handler(
|
|
guard, encoded(official_payload("PreToolUse")), fixture
|
|
)
|
|
audit_result = run_handler(
|
|
audit,
|
|
encoded(official_payload("PostToolUse", tool_response={"stdout": ""})),
|
|
fixture,
|
|
)
|
|
malformed = run_handler(audit, b"{ malformed", fixture)
|
|
|
|
assert guard_result.returncode == 0
|
|
assert b"Failed to resolve agent status" in guard_result.stderr
|
|
assert audit_result.returncode == 0
|
|
assert malformed.returncode == 0
|
|
log = (fixture.agent_dir / "error.log").read_text(encoding="utf-8")
|
|
assert "invalid JSON" in log
|
|
assert "refused" in log.lower() or "urlopen" in log.lower()
|