Files
crewAI/lib/crewai-tools/tests/tools/test_search_tools.py
alex-clawd 9325e2f6a4 fix: add path and URL validation to RAG tools (#5310)
* fix: add path and URL validation to RAG tools

Add validation utilities to prevent unauthorized file reads and SSRF
when RAG tools accept LLM-controlled paths/URLs at runtime.

Changes:
- New crewai_tools.utilities.safe_path module with validate_file_path(),
  validate_directory_path(), and validate_url()
- File paths validated against base directory (defaults to cwd).
  Resolves symlinks and ../ traversal. Rejects escape attempts.
- URLs validated: file:// blocked entirely. HTTP/HTTPS resolves DNS
  and blocks private/reserved IPs (10.x, 172.16-31.x, 192.168.x,
  127.x, 169.254.x, 0.0.0.0, ::1, fc00::/7).
- Validation applied in RagTool.add() — catches all RAG search tools
  (JSON, CSV, PDF, TXT, DOCX, MDX, Directory, etc.)
- Removed file:// scheme support from DataTypes.from_content()
- CREWAI_TOOLS_ALLOW_UNSAFE_PATHS=true env var for backward compat
- 27 tests covering traversal, symlinks, private IPs, cloud metadata,
  IPv6, escape hatch, and valid paths/URLs

* fix: validate path/URL keyword args in RagTool.add()

The original patch validated positional *args but left all keyword
arguments (path=, file_path=, directory_path=, url=, website=,
github_url=, youtube_url=) unvalidated, providing a trivial bypass
for both path-traversal and SSRF checks.

Applies validate_file_path() to path/file_path/directory_path kwargs
and validate_url() to url/website/github_url/youtube_url kwargs before
they reach the adapter. Adds a regression-test file covering all eight
kwarg vectors plus the two existing positional-arg checks.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: address CodeQL and review comments on RAG path/URL validation

- Replace insecure tempfile.mktemp() with inline symlink target in test
- Remove unused 'target' variable and unused tempfile import
- Narrow broad except Exception: pass to only catch urlparse errors;
  validate_url ValueError now propagates instead of being silently swallowed
- Fix ruff B904 (raise-without-from-inside-except) in safe_path.py
- Fix ruff B007 (unused loop variable 'family') in safe_path.py
- Use validate_directory_path in DirectorySearchTool.add() so the
  public utility is exercised in production code

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* style: fix ruff format + remaining lint issues

* fix: resolve mypy type errors in RAG path/URL validation

- Cast sockaddr[0] to str() to satisfy mypy (socket.getaddrinfo returns
  sockaddr where [0] is str but typed as str | int)
- Remove now-unnecessary `type: ignore[assignment]` and
  `type: ignore[literal-required]` comments in rag_tool.py

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: unroll dynamic TypedDict key loops to satisfy mypy literal-required

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* test: allow tmp paths in RAG data-type tests via CREWAI_TOOLS_ALLOW_UNSAFE_PATHS

TemporaryDirectory creates files under /tmp/ which is outside CWD and is
correctly blocked by the new path validation.  These tests exercise
data-type handling, not security, so add an autouse fixture that sets
CREWAI_TOOLS_ALLOW_UNSAFE_PATHS=true for the whole file.  Path/URL
security is covered by test_rag_tool_path_validation.py.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* test: allow tmp paths in search-tool and rag_tool tests via CREWAI_TOOLS_ALLOW_UNSAFE_PATHS

test_search_tools.py has tests for TXTSearchTool, CSVSearchTool,
MDXSearchTool, JSONSearchTool, and DirectorySearchTool that create
files under /tmp/ via tempfile, which is outside CWD and correctly
blocked by the new path validation.  rag_tool_test.py has one test
that calls tool.add() with a TemporaryDirectory path.

Add the same autouse allow_tmp_paths fixture used in
test_rag_tool_add_data_type.py.  Security is covered separately by
test_rag_tool_path_validation.py.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* chore: update tool specifications

* docs: document CodeInterpreterTool removal and RAG path/URL validation

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: address three review comments on path/URL validation

- safe_path._is_private_or_reserved: after unwrapping IPv4-mapped IPv6
  to IPv4, only check against IPv4 networks to avoid TypeError when
  comparing an IPv4Address against IPv6Network objects.
- safe_path.validate_file_path: handle filesystem-root base_dir ('/')
  by not appending os.sep when the base already ends with a separator,
  preventing the '//'-prefix bug.
- rag_tool.add: path-detection heuristic now checks for both '/' and
  os.sep so forward-slash paths are caught on Windows as well as Unix.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: remove unused _BLOCKED_NETWORKS variable after IPv4/IPv6 split

* chore: update tool specifications

---------

Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
2026-04-07 13:29:45 -03:00

364 lines
12 KiB
Python

import os
from pathlib import Path
import tempfile
from unittest.mock import MagicMock
from crewai_tools.rag.data_types import DataType
from crewai_tools.tools import (
CSVSearchTool,
CodeDocsSearchTool,
DOCXSearchTool,
DirectorySearchTool,
GithubSearchTool,
JSONSearchTool,
MDXSearchTool,
PDFSearchTool,
TXTSearchTool,
WebsiteSearchTool,
XMLSearchTool,
YoutubeChannelSearchTool,
YoutubeVideoSearchTool,
)
from crewai_tools.tools.rag.rag_tool import Adapter
import pytest
@pytest.fixture(autouse=True)
def allow_tmp_paths(monkeypatch: pytest.MonkeyPatch) -> None:
"""Allow absolute paths outside CWD (e.g. /tmp/) for these search-tool tests.
Path validation is tested separately in test_rag_tool_path_validation.py.
"""
monkeypatch.setenv("CREWAI_TOOLS_ALLOW_UNSAFE_PATHS", "true")
@pytest.fixture
def mock_adapter():
mock_adapter = MagicMock(spec=Adapter)
return mock_adapter
@pytest.mark.vcr()
def test_directory_search_tool():
with tempfile.TemporaryDirectory() as temp_dir:
test_file = Path(temp_dir) / "test.txt"
test_file.write_text("This is a test file for directory search")
tool = DirectorySearchTool(directory=temp_dir)
result = tool._run(search_query="test file")
assert "test file" in result.lower()
def test_pdf_search_tool(mock_adapter):
mock_adapter.query.return_value = "this is a test"
tool = PDFSearchTool(pdf="test.pdf", adapter=mock_adapter)
result = tool._run(query="test content")
assert "this is a test" in result.lower()
mock_adapter.add.assert_called_once_with("test.pdf", data_type=DataType.PDF_FILE)
mock_adapter.query.assert_called_once_with(
"test content", similarity_threshold=0.6, limit=5
)
mock_adapter.query.reset_mock()
mock_adapter.add.reset_mock()
tool = PDFSearchTool(adapter=mock_adapter)
result = tool._run(pdf="test.pdf", query="test content")
assert "this is a test" in result.lower()
mock_adapter.add.assert_called_once_with("test.pdf", data_type=DataType.PDF_FILE)
mock_adapter.query.assert_called_once_with(
"test content", similarity_threshold=0.6, limit=5
)
@pytest.mark.vcr()
def test_txt_search_tool():
with tempfile.NamedTemporaryFile(suffix=".txt", delete=False) as temp_file:
temp_file.write(b"This is a test file for txt search")
temp_file_path = temp_file.name
try:
tool = TXTSearchTool()
tool.add(temp_file_path)
result = tool._run(search_query="test file")
assert "test file" in result.lower()
finally:
os.unlink(temp_file_path)
def test_docx_search_tool(mock_adapter):
mock_adapter.query.return_value = "this is a test"
tool = DOCXSearchTool(docx="test.docx", adapter=mock_adapter)
result = tool._run(search_query="test content")
assert "this is a test" in result.lower()
mock_adapter.add.assert_called_once_with("test.docx", data_type=DataType.DOCX)
mock_adapter.query.assert_called_once_with(
"test content", similarity_threshold=0.6, limit=5
)
mock_adapter.query.reset_mock()
mock_adapter.add.reset_mock()
tool = DOCXSearchTool(adapter=mock_adapter)
result = tool._run(docx="test.docx", search_query="test content")
assert "this is a test" in result.lower()
mock_adapter.add.assert_called_once_with("test.docx", data_type=DataType.DOCX)
mock_adapter.query.assert_called_once_with(
"test content", similarity_threshold=0.6, limit=5
)
@pytest.mark.vcr()
def test_json_search_tool():
with tempfile.NamedTemporaryFile(suffix=".json", delete=False) as temp_file:
temp_file.write(b'{"test": "This is a test JSON file"}')
temp_file_path = temp_file.name
try:
tool = JSONSearchTool()
result = tool._run(search_query="test JSON", json_path=temp_file_path)
assert "test json" in result.lower()
finally:
os.unlink(temp_file_path)
def test_xml_search_tool(mock_adapter):
mock_adapter.query.return_value = "this is a test"
tool = XMLSearchTool(adapter=mock_adapter)
result = tool._run(search_query="test XML", xml="test.xml")
assert "this is a test" in result.lower()
mock_adapter.add.assert_called_once_with("test.xml")
mock_adapter.query.assert_called_once_with(
"test XML", similarity_threshold=0.6, limit=5
)
@pytest.mark.vcr()
def test_csv_search_tool():
with tempfile.NamedTemporaryFile(suffix=".csv", delete=False) as temp_file:
temp_file.write(b"name,description\ntest,This is a test CSV file")
temp_file_path = temp_file.name
try:
tool = CSVSearchTool()
tool.add(temp_file_path)
result = tool._run(search_query="test CSV")
assert "test csv" in result.lower()
finally:
os.unlink(temp_file_path)
@pytest.mark.vcr()
def test_mdx_search_tool():
with tempfile.NamedTemporaryFile(suffix=".mdx", delete=False) as temp_file:
temp_file.write(b"# Test MDX\nThis is a test MDX file")
temp_file_path = temp_file.name
try:
tool = MDXSearchTool()
tool.add(temp_file_path)
result = tool._run(search_query="test MDX")
assert "test mdx" in result.lower()
finally:
os.unlink(temp_file_path)
def test_website_search_tool(mock_adapter):
mock_adapter.query.return_value = "this is a test"
website = "https://crewai.com"
search_query = "what is crewai?"
tool = WebsiteSearchTool(website=website, adapter=mock_adapter)
result = tool._run(search_query=search_query)
mock_adapter.query.assert_called_once_with(
"what is crewai?", similarity_threshold=0.6, limit=5
)
mock_adapter.add.assert_called_once_with(website, data_type=DataType.WEBSITE)
assert "this is a test" in result.lower()
mock_adapter.query.reset_mock()
mock_adapter.add.reset_mock()
tool = WebsiteSearchTool(adapter=mock_adapter)
result = tool._run(website=website, search_query=search_query)
mock_adapter.query.assert_called_once_with(
"what is crewai?", similarity_threshold=0.6, limit=5
)
mock_adapter.add.assert_called_once_with(website, data_type=DataType.WEBSITE)
assert "this is a test" in result.lower()
def test_youtube_video_search_tool(mock_adapter):
mock_adapter.query.return_value = "some video description"
youtube_video_url = "https://www.youtube.com/watch?v=sample-video-id"
search_query = "what is the video about?"
tool = YoutubeVideoSearchTool(
youtube_video_url=youtube_video_url,
adapter=mock_adapter,
)
result = tool._run(search_query=search_query)
assert "some video description" in result
mock_adapter.add.assert_called_once_with(
youtube_video_url, data_type=DataType.YOUTUBE_VIDEO
)
mock_adapter.query.assert_called_once_with(
search_query, similarity_threshold=0.6, limit=5
)
mock_adapter.query.reset_mock()
mock_adapter.add.reset_mock()
tool = YoutubeVideoSearchTool(adapter=mock_adapter)
result = tool._run(youtube_video_url=youtube_video_url, search_query=search_query)
assert "some video description" in result
mock_adapter.add.assert_called_once_with(
youtube_video_url, data_type=DataType.YOUTUBE_VIDEO
)
mock_adapter.query.assert_called_once_with(
search_query, similarity_threshold=0.6, limit=5
)
def test_youtube_channel_search_tool(mock_adapter):
mock_adapter.query.return_value = "channel description"
youtube_channel_handle = "@crewai"
search_query = "what is the channel about?"
tool = YoutubeChannelSearchTool(
youtube_channel_handle=youtube_channel_handle, adapter=mock_adapter
)
result = tool._run(search_query=search_query)
assert "channel description" in result
mock_adapter.add.assert_called_once_with(
youtube_channel_handle, data_type=DataType.YOUTUBE_CHANNEL
)
mock_adapter.query.assert_called_once_with(
search_query, similarity_threshold=0.6, limit=5
)
mock_adapter.query.reset_mock()
mock_adapter.add.reset_mock()
tool = YoutubeChannelSearchTool(adapter=mock_adapter)
result = tool._run(
youtube_channel_handle=youtube_channel_handle, search_query=search_query
)
assert "channel description" in result
mock_adapter.add.assert_called_once_with(
youtube_channel_handle, data_type=DataType.YOUTUBE_CHANNEL
)
mock_adapter.query.assert_called_once_with(
search_query, similarity_threshold=0.6, limit=5
)
def test_code_docs_search_tool(mock_adapter):
mock_adapter.query.return_value = "test documentation"
docs_url = "https://crewai.com/any-docs-url"
search_query = "test documentation"
tool = CodeDocsSearchTool(docs_url=docs_url, adapter=mock_adapter)
result = tool._run(search_query=search_query)
assert "test documentation" in result
mock_adapter.add.assert_called_once_with(docs_url, data_type=DataType.DOCS_SITE)
mock_adapter.query.assert_called_once_with(
search_query, similarity_threshold=0.6, limit=5
)
mock_adapter.query.reset_mock()
mock_adapter.add.reset_mock()
tool = CodeDocsSearchTool(adapter=mock_adapter)
result = tool._run(docs_url=docs_url, search_query=search_query)
assert "test documentation" in result
mock_adapter.add.assert_called_once_with(docs_url, data_type=DataType.DOCS_SITE)
mock_adapter.query.assert_called_once_with(
search_query, similarity_threshold=0.6, limit=5
)
def test_github_search_tool(mock_adapter):
mock_adapter.query.return_value = "repo description"
# ensure the provided repo and content types are used after initialization
tool = GithubSearchTool(
gh_token="test_token",
github_repo="crewai/crewai",
content_types=["code"],
adapter=mock_adapter,
)
result = tool._run(search_query="tell me about crewai repo")
assert "repo description" in result
mock_adapter.add.assert_called_once_with(
"https://github.com/crewai/crewai",
data_type=DataType.GITHUB,
metadata={"content_types": ["code"], "gh_token": "test_token"},
)
mock_adapter.query.assert_called_once_with(
"tell me about crewai repo", similarity_threshold=0.6, limit=5
)
# ensure content types provided by run call is used
mock_adapter.query.reset_mock()
mock_adapter.add.reset_mock()
tool = GithubSearchTool(gh_token="test_token", adapter=mock_adapter)
result = tool._run(
github_repo="crewai/crewai",
content_types=["code", "issue"],
search_query="tell me about crewai repo",
)
assert "repo description" in result
mock_adapter.add.assert_called_once_with(
"https://github.com/crewai/crewai",
data_type=DataType.GITHUB,
metadata={"content_types": ["code", "issue"], "gh_token": "test_token"},
)
mock_adapter.query.assert_called_once_with(
"tell me about crewai repo", similarity_threshold=0.6, limit=5
)
# ensure default content types are used if not provided
mock_adapter.query.reset_mock()
mock_adapter.add.reset_mock()
tool = GithubSearchTool(gh_token="test_token", adapter=mock_adapter)
result = tool._run(
github_repo="crewai/crewai",
search_query="tell me about crewai repo",
)
assert "repo description" in result
mock_adapter.add.assert_called_once_with(
"https://github.com/crewai/crewai",
data_type=DataType.GITHUB,
metadata={
"content_types": ["code", "repo", "pr", "issue"],
"gh_token": "test_token",
},
)
mock_adapter.query.assert_called_once_with(
"tell me about crewai repo", similarity_threshold=0.6, limit=5
)
# ensure nothing is added if no repo is provided
mock_adapter.query.reset_mock()
mock_adapter.add.reset_mock()
tool = GithubSearchTool(gh_token="test_token", adapter=mock_adapter)
result = tool._run(search_query="tell me about crewai repo")
mock_adapter.add.assert_not_called()
mock_adapter.query.assert_called_once_with(
"tell me about crewai repo", similarity_threshold=0.6, limit=5
)