cross-validated-search

pip install zero-api-key-web-search

# Search the web — no API key needed
zero-search "Python 3.13 release" --json

# Build citation-ready LLM context
zero-context "Python 3.13 stable release" --goggles docs-first

# Read a page
zero-browse "https://docs.python.org/3/whatsnew/" --json

# Verify a claim
zero-verify "Python 3.13 is the latest stable release" --deep --json

# Full evidence report
zero-report "Python 3.13 stable release" \
  --claim "Python 3.13 is the latest stable release" --deep --json

zero-search "Python 3.13 release" --json

# Start SearXNG locally
./scripts/start-searxng.sh
export ZERO_SEARCH_SEARXNG_URL="http://127.0.0.1:8080"

zero-search "AI regulation" --profile free-verified --json

cp .env.searxng.example .env
docker compose -f docker-compose.searxng.yml up -d

# Interactive setup wizard — tests your key and zone
zero-setup

# Or set environment variables directly
export ZERO_SEARCH_BRIGHTDATA_API_KEY="your-key"
export ZERO_SEARCH_BRIGHTDATA_ZONE="serp_api1"

# Search across different engines
zero-search "news" --provider brightdata --engine google --type news --region us-en --json
zero-search "news" --provider brightdata --engine bing --type news --region gb-en --json
zero-search "news" --provider brightdata --engine yandex --region ru-ru --json

# Setup wizard handles zone creation guidance
zero-setup

# Or set the zone manually
export ZERO_SEARCH_BRIGHTDATA_API_KEY="your-key"
export ZERO_SEARCH_BRIGHTDATA_ZONE="serp_api1"
export ZERO_SEARCH_BRIGHTDATA_UNLOCKER_ZONE="web_unlocker1"

# Browse automatically falls back to Web Unlocker on 403/429
zero-browse "https://protected-site.com/article" --json

# Or force Web Unlocker
zero-browse "https://protected-site.com/article" --use-unlocker always --json

export ZERO_SEARCH_BRIGHTDATA_API_KEY="your-key"
export ZERO_SEARCH_SEARXNG_URL="http://127.0.0.1:8080"

zero-report "AI regulation news" --profile max-evidence --deep --json

zero-search "FastAPI lifespan docs" --profile free-verified --goggles docs-first
zero-context "FastAPI lifespan docs" --profile free --goggles docs-first
zero-report "AI regulation news" --profile production-unlock --json

{
  "mcpServers": {
    "zero-api-key-web-search": {
      "command": "zero-mcp"
    }
  }
}

{
  "mcpServers": {
    "zero-api-key-web-search": {
      "command": "npx",
      "args": ["zero-api-key-web-search", "zero-mcp"]
    }
  }
}

# Interactive wizard
zero-setup

# Quick status check
zero-setup --status

# Test a Bright Data API key
zero-setup --test-brightdata YOUR_API_KEY

# Test a Bright Data Web Unlocker zone
zero-setup --test-brightdata-unlocker YOUR_API_KEY --unlocker-zone web_unlocker1

# Test a SearXNG instance
zero-setup --test-searxng http://localhost:8080

from zero_api_key_web_search.providers import BrightDataProvider

provider = BrightDataProvider(api_key="your-key", zone="serp_api1")

# Google (default)
results = provider.search("Python 3.13", search_type="text", region="us-en")

# Bing, Yandex, Baidu, Yahoo, Naver, DuckDuckGo
results = provider.search("AI regulation", search_type="news", engine="bing")

# Markdown output for LLM consumption
results = provider.search("climate change", data_format="markdown")

from zero_api_key_web_search.providers import WebUnlockerProvider

provider = WebUnlockerProvider(api_key="your-key", zone="web_unlocker1")

# Get page content as markdown
result = provider.unlock("https://protected-site.com/article", data_format="markdown")
print(result["content"])  # Clean markdown of the page

# With country targeting
result = provider.unlock("https://geo-restricted.com", country="us")

from zero_api_key_web_search.browse_page import browse

# Auto-fallback (default) — try direct, then unlocker on 403/429
result = browse("https://protected-site.com/article")

# Always use Web Unlocker
result = browse("https://protected-site.com/article", use_unlocker=True)

# Never use Web Unlocker
result = browse("https://protected-site.com/article", use_unlocker=False)

zero_api_key_web_search/
  core.py              # UltimateSearcher — search, verify, report engine
  browse_page.py       # Readability-style page extraction + Web Unlocker fallback
  mcp_server.py        # MCP server (8 tools)
  provider_setup.py    # Interactive setup wizard (zero-setup)
  cache.py             # Response caching
  transport.py         # SSL/TLS helpers
  search_web.py        # CLI: zero-search
  context.py            # CLI: zero-context
  verify_claim.py       # CLI: zero-verify
  evidence_report.py    # CLI: zero-report
  providers/
    base.py             # SearchProvider protocol (sync + async)
    ddgs.py             # DuckDuckGo provider (free, zero-config)
    searxng.py          # SearXNG provider (free, self-hosted)
    brightdata.py        # Bright Data SERP — 7 engines, markdown, AI Overviews
    web_unlocker.py      # Bright Data Web Unlocker — blocked/CAPTCHA/geo pages
  skills/
    SKILL.md            # Bundled OpenClaw skill

pip install zero-api-key-web-search

pip install -e ".[dev]"
python -m pytest tests/ -q           # 98 tests
ruff check zero_api_key_web_search/ tests/
pyright zero_api_key_web_search/     # 0 errors
coverage report --fail-under=80       # 85% coverage

{
  "verdict": "contested",
  "confidence": "MEDIUM",
  "executive_summary": "Evidence is split...",
  "verdict_rationale": ["Source A supports...", "Source B contradicts..."],
  "coverage_warnings": ["Single-provider evidence path."],
  "source_digest": [
    {"title": "...", "url": "...", "classification": "supporting", "evidence_strength": 3}
  ],
  "next_steps": ["Add a second provider for cross-validation."]
}