Files
odysseus/tests/test_preview_execution_evidence.py
T

77 lines
3.5 KiB
Python

"""Execution outcomes must reach the model, not just the tool-event UI."""
import json
import pytest
from src.agent_tools.subprocess_tools import BashTool
from src.clean_agent_preview import preview_tool_result_text
@pytest.mark.asyncio
async def test_failed_shell_retains_exit_status_and_both_streams_for_followup(tmp_path):
from src.tool_execution import _active_workspace
token = _active_workspace.set(str(tmp_path))
try:
result = await BashTool().execute("printf 'PHASE_ONE_DONE\\n'; printf 'CHECK_FAILED\\n' >&2; exit 7", {})
finally:
_active_workspace.reset(token)
assert result['exit_code'] == 7
observed = preview_tool_result_text(result, 'bash', {})
assert 'PHASE_ONE_DONE' in observed and 'CHECK_FAILED' in observed
assert json.loads(observed)['exit_code'] == 7
def test_partial_output_does_not_hide_error_or_timeout_evidence():
result = {'output': 'partial data', 'error': 'Timed out', 'exit_code': 124,
'stdout': 'first chunk', 'stderr': 'diagnostic'}
observed = json.loads(preview_tool_result_text(result, 'bash', {}))
assert observed == result
def test_failure_status_survives_observation_truncation():
observed = preview_tool_result_text({'output': 'x' * 10000, 'exit_code': 9, 'error': 'Incomplete scan'}, 'grep', {})
assert observed.startswith('{"exit_code": 9, "error": "Incomplete scan"')
assert 'truncated' in observed
assert len(observed) < 8100
def test_successful_plain_output_keeps_existing_compact_representation():
assert preview_tool_result_text({'output': 'alpha', 'exit_code': 0}, 'read_file', {}) == 'alpha'
@pytest.mark.asyncio
async def test_unoffered_proposal_is_marked_blocked_not_an_execution_attempt(monkeypatch):
import src.clean_agent_preview as preview
from src.tool_policy import ToolPolicy
from src.turn_contract import resolve_full_inventory_contract
responses = iter([
{'tool_calls': [{'index': 0, 'id': 'unoffered', 'function': {'name': 'manage_skills',
'arguments': json.dumps({'action': 'view', 'name': 'not-executed'})}}]},
{'content': 'That tool is unavailable.'},
])
class Response:
async def __aenter__(self): return self
async def __aexit__(self, *args): pass
def raise_for_status(self): pass
async def aiter_lines(self):
yield 'data: ' + json.dumps({'choices': [{'delta': next(responses)}]})
yield 'data: [DONE]'
class Client:
def __init__(self, **kwargs): pass
async def __aenter__(self): return self
async def __aexit__(self, *args): pass
def stream(self, *args, **kwargs): return Response()
monkeypatch.setattr(preview.httpx, 'AsyncClient', Client)
policy = ToolPolicy()
contract = resolve_full_inventory_contract(schemas=[], policy=policy)
raw = [chunk async for chunk in preview.stream_preview(endpoint_url='http://test', model='test',
messages=[{'role': 'user', 'content': 'Show that skill'}], headers={}, turn_contract=contract,
session_id='unoffered-test', owner='test', disabled_tools=set(), tool_policy=policy)]
events = [json.loads(chunk[6:]) for chunk in raw if '[DONE]' not in chunk]
assert not any(e.get('type') == 'tool_start' for e in events)
output = next(e for e in events if e.get('type') == 'tool_output')
assert output['execution_attempted'] is False
assert output['blocked'] is True
saved = next(e['data']['tool_events'] for e in events if e.get('type') == 'metrics')
assert saved == [output]