SEO Intel

Local-first competitive SEO intelligence. Point it at your site + competitors, get keyword gaps, content audits, and visual dashboards. All data stays on your machine.

Crawl → Extract (local AI) → Analyze (cloud AI) → Dashboard

Your site + competitors (Playwright crawler)
    ↓ structured extraction
Qwen 3.5 via Ollama (local, free)
    ↓ stored in
SQLite database (WAL mode)
    ↓ competitive analysis
Gemini / Claude / GPT (your API key)
    ↓ visual reports
Self-contained HTML dashboards (Chart.js)

Quick Start

# Install globally
npm install -g seo-intel

# Run the setup wizard (auto-detects OpenClaw for agent-powered setup)
seo-intel setup

The setup wizard handles everything: dependency checks, model selection, API keys, project configuration, and pipeline validation.

Requirements

• Node.js 22.5+ (uses built-in SQLite)
• Ollama with a Qwen model (for local extraction)
• One API key for analysis: Gemini (recommended), Claude, OpenAI, or DeepSeek

Manual Setup

npm install -g seo-intel
seo-intel setup --classic    # traditional CLI wizard
# or
seo-intel setup              # agent-powered if OpenClaw is running

Usage

# Full pipeline
seo-intel crawl myproject       # crawl target + competitors
seo-intel extract myproject     # local AI extraction (Ollama)
seo-intel analyze myproject     # competitive gap analysis
seo-intel html myproject        # generate dashboard
seo-intel serve                 # open dashboard at localhost:3000

# Agentic exports — turn data into implementation briefs
seo-intel export-actions myproject --scope technical   # free: broken links, missing schemas, orphans
seo-intel export-actions myproject --scope all         # full: technical + competitive + suggestive
seo-intel competitive-actions myproject --vs rival.com # what competitors have that you don't
seo-intel suggest-usecases myproject --scope docs      # infer what pages/docs should exist

Commands

Free

| Command | Description | |---------|-------------| | setup | First-time wizard — auto-detects OpenClaw for agent-powered setup | | crawl <project> | Crawl target + competitor sites | | status | System status, crawl freshness, license info | | html <project> | Generate crawl-only dashboard | | serve | Start local dashboard server (port 3000) | | export-actions <project> --scope technical | Technical SEO audit from crawl data | | schemas <project> | Schema.org coverage analysis | | update | Check for updates |

Solo (€19.99/mo · ukkometa.fi/seo-intel)

| Command | Description | |---------|-------------| | extract <project> | Local AI extraction via Ollama | | analyze <project> | Full competitive gap analysis | | export-actions <project> | All export scopes (technical + competitive + suggestive) | | competitive-actions <project> | Competitive gap export with --vs domain filter | | suggest-usecases <project> | Infer what pages/features to build from competitor data | | keywords <project> | Keyword gap matrix | | run <project> | Full pipeline: crawl → extract → analyze → report | | brief <project> | AI content briefs for gap topics | | velocity <project> | Content publishing velocity tracker | | shallow <project> | Find "shallow champion" pages to outrank | | decay <project> | Find stale competitor content | | headings-audit <project> | H1-H6 structure analysis | | orphans <project> | Orphaned pages detection | | entities <project> | Entity coverage gap analysis | | friction <project> | Conversion friction detection | | js-delta <project> | JS-rendered vs raw HTML comparison |

Project Configuration

Create a project config in config/:

{
  "project": "myproject",
  "context": {
    "siteName": "My Site",
    "url": "https://example.com",
    "industry": "Your industry description",
    "audience": "Your target audience",
    "goal": "Your SEO objective"
  },
  "target": {
    "domain": "example.com",
    "maxPages": 200,
    "crawlMode": "standard"
  },
  "competitors": [
    { "domain": "competitor1.com", "maxPages": 100 },
    { "domain": "competitor2.com", "maxPages": 100 }
  ]
}

Or use the setup wizard: seo-intel setup

Managing Competitors

seo-intel competitors myproject                    # list all
seo-intel competitors myproject --add new-rival.com
seo-intel competitors myproject --remove old-rival.com

Web Setup Wizard

seo-intel serve
# Open http://localhost:3000/setup

The 6-step web wizard guides you through:

1. System Check — Node, Ollama, Playwright, GPU detection
2. Models — VRAM-based model recommendations
3. Project — Target domain + competitors
4. Search Console — CSV upload or OAuth API
5. Pipeline Test — Validates the full pipeline
6. Done — Your first CLI commands

If OpenClaw is running, you'll see an option for agent-powered setup that handles everything conversationally — including troubleshooting, dependency installation, and OAuth configuration.

Model Configuration

Extraction (local, free)

SEO Intel uses Ollama for local AI extraction. Edit .env:

OLLAMA_URL=http://localhost:11434
OLLAMA_MODEL=qwen3.5:9b        # recommended (needs 6GB+ VRAM)
OLLAMA_CTX=16384

Model recommendations by VRAM:

• 3-4 GB → qwen3.5:4b
• 6-8 GB → qwen3.5:9b (recommended)
• 16+ GB → qwen3.5:27b

Analysis (cloud, user's API key)

You need at least one API key in .env:

GEMINI_API_KEY=your-key          # recommended (~$0.01/analysis)
# or
ANTHROPIC_API_KEY=your-key       # highest quality
# or
OPENAI_API_KEY=your-key          # solid all-around
# or
DEEPSEEK_API_KEY=your-key        # budget option

Google Search Console

Upload your GSC data for ranking insights:

1. Go to Google Search Console
2. Export Performance data as CSV
3. Upload via the web wizard or place CSVs in gsc/<project>/

License

Free Tier

• 1 project, 500 pages/domain
• Crawl, extract, setup, basic reports

Solo (€19.99/mo · €199.99/yr)

• Unlimited projects and pages
• All analysis commands, GSC insights, scheduling

# Set your license key
echo "SEO_INTEL_LICENSE=SI-xxxx-xxxx-xxxx-xxxx" >> .env

Get a key at ukkometa.fi/seo-intel

Updates

seo-intel update              # check for updates
seo-intel update --apply      # auto-apply via npm

Updates are checked automatically in the background and shown at the end of seo-intel status.

Security

• All data stays local — no telemetry, no cloud sync
• Scraped content is HTML-stripped and sanitized before reaching any model
• Extraction outputs are validated against schema before DB insert
• API keys are stored in .env (gitignored)
• OAuth tokens stored in .tokens/ (gitignored)

OpenClaw Integration

If you have OpenClaw installed:

seo-intel setup              # auto-detects gateway, uses agent
seo-intel setup --agent      # require agent setup
seo-intel setup --classic    # force manual wizard

The OpenClaw agent provides:

• Conversational setup with real-time troubleshooting
• Automatic dependency installation
• Smart model recommendations
• Security update notifications

Built by ukkometa.fi — local-first SEO intelligence.