Files
deerflow-factory/deer-flow/backend/tests/test_credential_loader.py
DATA 6de0bf9f5b Initial commit: hardened DeerFlow factory
Vendored deer-flow upstream (bytedance/deer-flow) plus prompt-injection
hardening:

- New deerflow.security package: content_delimiter, html_cleaner,
  sanitizer (8 layers — invisible chars, control chars, symbols, NFC,
  PUA, tag chars, horizontal whitespace collapse with newline/tab
  preservation, length cap)
- New deerflow.community.searx package: web_search, web_fetch,
  image_search backed by a private SearX instance, every external
  string sanitized and wrapped in <<<EXTERNAL_UNTRUSTED_CONTENT>>>
  delimiters
- All native community web providers (ddg_search, tavily, exa,
  firecrawl, jina_ai, infoquest, image_search) replaced with hard-fail
  stubs that raise NativeWebToolDisabledError at import time, so a
  misconfigured tool.use path fails loud rather than silently falling
  back to unsanitized output
- Native client back-doors (jina_client.py, infoquest_client.py)
  stubbed too
- Native-tool tests quarantined under tests/_disabled_native/
  (collect_ignore_glob via local conftest.py)
- Sanitizer Layer 7 fix: only collapse horizontal whitespace, preserve
  newlines and tabs so list/table structure survives
- Hardened runtime config.yaml references only the searx-backed tools
- Factory overlay (backend/) kept in sync with deer-flow tree as a
  reference / source

See HARDENING.md for the full audit trail and verification steps.
2026-04-12 14:23:57 +02:00

157 lines
4.7 KiB
Python

import json
import os
from deerflow.models.credential_loader import (
load_claude_code_credential,
load_codex_cli_credential,
)
def _clear_claude_code_env(monkeypatch) -> None:
for env_var in (
"CLAUDE_CODE_OAUTH_TOKEN",
"ANTHROPIC_AUTH_TOKEN",
"CLAUDE_CODE_OAUTH_TOKEN_FILE_DESCRIPTOR",
"CLAUDE_CODE_CREDENTIALS_PATH",
):
monkeypatch.delenv(env_var, raising=False)
def test_load_claude_code_credential_from_direct_env(monkeypatch):
_clear_claude_code_env(monkeypatch)
monkeypatch.setenv("CLAUDE_CODE_OAUTH_TOKEN", " sk-ant-oat01-env ")
cred = load_claude_code_credential()
assert cred is not None
assert cred.access_token == "sk-ant-oat01-env"
assert cred.refresh_token == ""
assert cred.source == "claude-cli-env"
def test_load_claude_code_credential_from_anthropic_auth_env(monkeypatch):
_clear_claude_code_env(monkeypatch)
monkeypatch.setenv("ANTHROPIC_AUTH_TOKEN", "sk-ant-oat01-anthropic-auth")
cred = load_claude_code_credential()
assert cred is not None
assert cred.access_token == "sk-ant-oat01-anthropic-auth"
assert cred.source == "claude-cli-env"
def test_load_claude_code_credential_from_file_descriptor(monkeypatch):
_clear_claude_code_env(monkeypatch)
read_fd, write_fd = os.pipe()
try:
os.write(write_fd, b"sk-ant-oat01-fd")
os.close(write_fd)
monkeypatch.setenv("CLAUDE_CODE_OAUTH_TOKEN_FILE_DESCRIPTOR", str(read_fd))
cred = load_claude_code_credential()
finally:
os.close(read_fd)
assert cred is not None
assert cred.access_token == "sk-ant-oat01-fd"
assert cred.refresh_token == ""
assert cred.source == "claude-cli-fd"
def test_load_claude_code_credential_from_override_path(tmp_path, monkeypatch):
_clear_claude_code_env(monkeypatch)
cred_path = tmp_path / "claude-credentials.json"
cred_path.write_text(
json.dumps(
{
"claudeAiOauth": {
"accessToken": "sk-ant-oat01-test",
"refreshToken": "sk-ant-ort01-test",
"expiresAt": 4_102_444_800_000,
}
}
)
)
monkeypatch.setenv("CLAUDE_CODE_CREDENTIALS_PATH", str(cred_path))
cred = load_claude_code_credential()
assert cred is not None
assert cred.access_token == "sk-ant-oat01-test"
assert cred.refresh_token == "sk-ant-ort01-test"
assert cred.source == "claude-cli-file"
def test_load_claude_code_credential_ignores_directory_path(tmp_path, monkeypatch):
_clear_claude_code_env(monkeypatch)
cred_dir = tmp_path / "claude-creds-dir"
cred_dir.mkdir()
monkeypatch.setenv("CLAUDE_CODE_CREDENTIALS_PATH", str(cred_dir))
assert load_claude_code_credential() is None
def test_load_claude_code_credential_falls_back_to_default_file_when_override_is_invalid(tmp_path, monkeypatch):
_clear_claude_code_env(monkeypatch)
monkeypatch.setenv("HOME", str(tmp_path))
cred_dir = tmp_path / "claude-creds-dir"
cred_dir.mkdir()
monkeypatch.setenv("CLAUDE_CODE_CREDENTIALS_PATH", str(cred_dir))
default_path = tmp_path / ".claude" / ".credentials.json"
default_path.parent.mkdir()
default_path.write_text(
json.dumps(
{
"claudeAiOauth": {
"accessToken": "sk-ant-oat01-default",
"refreshToken": "sk-ant-ort01-default",
"expiresAt": 4_102_444_800_000,
}
}
)
)
cred = load_claude_code_credential()
assert cred is not None
assert cred.access_token == "sk-ant-oat01-default"
assert cred.refresh_token == "sk-ant-ort01-default"
assert cred.source == "claude-cli-file"
def test_load_codex_cli_credential_supports_nested_tokens_shape(tmp_path, monkeypatch):
auth_path = tmp_path / "auth.json"
auth_path.write_text(
json.dumps(
{
"tokens": {
"access_token": "codex-access-token",
"account_id": "acct_123",
}
}
)
)
monkeypatch.setenv("CODEX_AUTH_PATH", str(auth_path))
cred = load_codex_cli_credential()
assert cred is not None
assert cred.access_token == "codex-access-token"
assert cred.account_id == "acct_123"
assert cred.source == "codex-cli"
def test_load_codex_cli_credential_supports_legacy_top_level_shape(tmp_path, monkeypatch):
auth_path = tmp_path / "auth.json"
auth_path.write_text(json.dumps({"access_token": "legacy-access-token"}))
monkeypatch.setenv("CODEX_AUTH_PATH", str(auth_path))
cred = load_codex_cli_credential()
assert cred is not None
assert cred.access_token == "legacy-access-token"
assert cred.account_id == ""