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.
This commit is contained in:
2026-04-12 14:23:57 +02:00
commit 6de0bf9f5b
889 changed files with 173052 additions and 0 deletions

163
deer-flow/scripts/check.py Normal file
View File

@@ -0,0 +1,163 @@
#!/usr/bin/env python3
"""Cross-platform dependency checker for DeerFlow."""
from __future__ import annotations
import shutil
import subprocess
import sys
def configure_stdio() -> None:
"""Prefer UTF-8 output so Unicode status markers render on Windows."""
for stream_name in ("stdout", "stderr"):
stream = getattr(sys, stream_name, None)
if hasattr(stream, "reconfigure"):
try:
stream.reconfigure(encoding="utf-8", errors="replace")
except (OSError, ValueError):
continue
def run_command(command: list[str]) -> str | None:
"""Run a command and return trimmed stdout, or None on failure."""
try:
result = subprocess.run(command, capture_output=True, text=True, check=True, shell=False)
except (OSError, subprocess.CalledProcessError):
return None
return result.stdout.strip() or result.stderr.strip()
def find_pnpm_command() -> list[str] | None:
"""Return a pnpm-compatible command that exists on this machine."""
candidates = [["pnpm"], ["pnpm.cmd"]]
if shutil.which("corepack"):
candidates.append(["corepack", "pnpm"])
for command in candidates:
if shutil.which(command[0]):
return command
return None
def parse_node_major(version_text: str) -> int | None:
version = version_text.strip()
if version.startswith("v"):
version = version[1:]
major_str = version.split(".", 1)[0]
if not major_str.isdigit():
return None
return int(major_str)
def main() -> int:
configure_stdio()
print("==========================================")
print(" Checking Required Dependencies")
print("==========================================")
print()
failed = False
print("Checking Node.js...")
node_path = shutil.which("node")
if node_path:
node_version = run_command(["node", "-v"])
if node_version:
major = parse_node_major(node_version)
if major is not None and major >= 22:
print(f" OK Node.js {node_version.lstrip('v')} (>= 22 required)")
else:
print(
f" FAIL Node.js {node_version.lstrip('v')} found, but version 22+ is required"
)
print(" Install from: https://nodejs.org/")
failed = True
else:
print(" INFO Unable to determine Node.js version")
print(" Install from: https://nodejs.org/")
failed = True
else:
print(" FAIL Node.js not found (version 22+ required)")
print(" Install from: https://nodejs.org/")
failed = True
print()
print("Checking pnpm...")
pnpm_command = find_pnpm_command()
if pnpm_command:
pnpm_version = run_command([*pnpm_command, "-v"])
if pnpm_version:
if pnpm_command[0] == "corepack":
print(f" OK pnpm {pnpm_version} (via Corepack)")
else:
print(f" OK pnpm {pnpm_version}")
else:
print(" INFO Unable to determine pnpm version")
failed = True
else:
print(" FAIL pnpm not found")
print(" Install: npm install -g pnpm")
print(" Or enable Corepack: corepack enable")
print(" Or visit: https://pnpm.io/installation")
failed = True
print()
print("Checking uv...")
if shutil.which("uv"):
uv_version_text = run_command(["uv", "--version"])
if uv_version_text:
uv_version_parts = uv_version_text.split()
uv_version = uv_version_parts[1] if len(uv_version_parts) > 1 else uv_version_text
print(f" OK uv {uv_version}")
else:
print(" INFO Unable to determine uv version")
failed = True
else:
print(" FAIL uv not found")
print(" Visit the official installation guide for your platform:")
print(" https://docs.astral.sh/uv/getting-started/installation/")
failed = True
print()
print("Checking nginx...")
if shutil.which("nginx"):
nginx_version_text = run_command(["nginx", "-v"])
if nginx_version_text and "/" in nginx_version_text:
nginx_version = nginx_version_text.split("/", 1)[1]
print(f" OK nginx {nginx_version}")
else:
print(" INFO nginx (version unknown)")
else:
print(" FAIL nginx not found")
print(" macOS: brew install nginx")
print(" Ubuntu: sudo apt install nginx")
print(" Windows: use WSL for local mode or use Docker mode")
print(" Or visit: https://nginx.org/en/download.html")
failed = True
print()
if not failed:
print("==========================================")
print(" OK All dependencies are installed!")
print("==========================================")
print()
print("You can now run:")
print(" make install - Install project dependencies")
print(" make setup - Create a minimal working config (recommended)")
print(" make config - Copy the full config template (manual setup)")
print(" make doctor - Verify config and dependency health")
print(" make dev - Start development server")
print(" make start - Start production server")
return 0
print("==========================================")
print(" FAIL Some dependencies are missing")
print("==========================================")
print()
print("Please install the missing tools and run 'make check' again.")
return 1
if __name__ == "__main__":
sys.exit(main())