Use Claude Code, Codex, VSCode, Pi, and OpenCode (and 6 other harnesses) for free (1.3B+ free tokens) from your terminal, app, IDE, or phone, and now from the browser with native browser sessions (multi-harness + multi-model) like OpenClaw (voice supported + ToS friendly)
Python
"""Fresh-process contracts for the work excluded from HTTP/tray startup."""
import subprocess
import sys
import pytest
@pytest.mark.parametrize(
("module", "forbidden"),
[
(
"free_claude_code.cli.desktop",
("uvicorn", "fastapi", "openai", "copilot", "tiktoken"),
),
(
"free_claude_code.runtime.bootstrap",
("openai", "copilot", "tiktoken", "telegram", "discord"),
),
],
)
def test_entrypoint_import_does_not_load_optional_runtime(module, forbidden):
script = f"""
import importlib
import sys
importlib.import_module({module!r})
assert not set({forbidden!r}).intersection(sys.modules), sorted(set({forbidden!r}).intersection(sys.modules))
"""
result = subprocess.run(
[sys.executable, "-c", script], capture_output=True, text=True, timeout=15
)
assert result.returncode == 0, result.stderr
def test_sdk_preparation_runs_in_worker_and_provider_construction_runs_on_owner_loop():
script = """
import asyncio
import sys
import threading
from unittest.mock import MagicMock
from free_claude_code.config.settings import Settings
from free_claude_code.providers.base import BaseProvider
from free_claude_code.providers.runtime.runtime import create_provider
from free_claude_code.providers.admission_policy import ProviderAdmissionLimits
from free_claude_code.providers.admission_registry import ProviderAdmissionRegistry
async def main():
loop = asyncio.get_running_loop()
thread = threading.get_ident()
provider = MagicMock(spec=BaseProvider)
def load():
assert threading.get_ident() != thread
assert "openai.resources" in sys.modules
def construct(*args):
assert asyncio.get_running_loop() is loop
assert threading.get_ident() == thread
return provider
return construct
assert await create_provider("groq", Settings(groq_api_keys="test"), ProviderAdmissionRegistry(ProviderAdmissionLimits(1, 2, 2)), provider_loaders={"groq": load}) is provider
await provider.cleanup()
asyncio.run(main())
"""
result = subprocess.run(
[sys.executable, "-c", script], capture_output=True, text=True, timeout=15
)
assert result.returncode == 0, result.stderr
def test_tokenizer_initializes_offline_with_an_empty_cache():
script = """
import os
import tempfile
import requests
def network_forbidden(*args, **kwargs):
raise AssertionError("tokenizer attempted a network request")
requests.get = network_forbidden
with tempfile.TemporaryDirectory() as cache:
os.environ["TIKTOKEN_CACHE_DIR"] = cache
from free_claude_code.core.token_estimation import initialize_token_estimation, estimate_text_tokens
encoder = initialize_token_estimation()
assert encoder is not None
assert estimate_text_tokens("hello world") == 2
assert estimate_text_tokens("abcdefgh") == 1
"""
result = subprocess.run(
[sys.executable, "-c", script], capture_output=True, text=True, timeout=15
)
assert result.returncode == 0, result.stderr