search-intelligence-skill

Use search-intelligence-skill to give any AI agent the ability to search the entire internet like an expert OSINT analyst, SEO engineer, and security researcher combined. All searches flow through your SearXNG instance — zero API keys, full privacy, 90+ engines.

The skill generates optimized dork queries, selects intelligent multi-step search strategies, translates operators across engines, routes queries to the best SearXNG engines, scores results by multi-signal relevance, and learns from results to refine searches automatically.

Setup (once)

Install the package

# From source (recommended)
git clone https://github.com/mouaad-ops/search-intelligence-skill.git
cd search-intelligence-skill
pip install -e .

# Or direct pip
pip install search-intelligence-skill # NOT yet working

Start a SearXNG instance (if you don't have one)

# Docker (quickest)
docker run -d \
  --name searxng \
  -p 8888:8080 \
  -e SEARXNG_SECRET=your-secret-key \
  searxng/searxng:latest

# Verify it's running
curl http://localhost:8888/healthz

Enable JSON API in SearXNG settings

# In searxng/settings.yml — ensure search formats include json
search:
  formats:
    - html
    - json

Initialize in code

from search_intelligence_skill import SearchSkill

# Default — localhost:8888
skill = SearchSkill()

# Custom instance
skill = SearchSkill(
    searxng_url="http://localhost:8888",
    timeout=30.0,
    max_retries=2,
    rate_limit=0.5,
    verify_ssl=True,
    auto_refine=True,
    max_refine_rounds=1,
)

# Verify connection
if skill.health_check():
    print("✓ SearXNG is reachable")
else:
    print("✗ Cannot reach SearXNG — check URL and port")

Common Commands

Natural language search (the main interface)

from search_intelligence_skill import SearchSkill

skill = SearchSkill(searxng_url="http://localhost:8888")

# Just describe what you want — the skill handles everything:
# intent detection, dork generation, engine selection, scoring
report = skill.search("find exposed .env files on example.com")

# Print LLM-ready formatted output
print(report.to_context())

# Access structured results
for r in report.top(5):
    print(f"[{r.relevance:.1f}] {r.title}")
    print(f"    {r.url}")
    print(f"    {r.snippet[:200]}")

Control search depth

from search_intelligence_skill import Depth

# Quick — 1-2 queries, single step, fast lookup
report = skill.search("what is CORS", depth="quick")

# Standard — 3-6 queries, multi-engine, good default
report = skill.search("python async frameworks comparison", depth="standard")

# Deep — 6-12 queries, multi-step strategies, thorough research
report = skill.search("security audit of target.com", depth="deep")

# Exhaustive — 12+ queries, full OSINT chains, complete sweep
report = skill.search("full recon on suspect-domain.com", depth="exhaustive")