MCP Server & CLI

1. Get an API key

Go to your dashboard, find the API Keys section and create one. Keys start with wr_live_.

2. MCP Server

The MCP server connects your AI assistant (Claude Desktop, Claude Code, Codex, Gemini...) directly to Web Resurrect. The agent can create a project, scrape, rewrite, generate images and publish to WordPress on its own.

Claude Desktop

Add this to your claude_desktop_config.json:

{
  "mcpServers": {
    "web-resurrect": {
      "command": "npx",
      "args": ["-y", "@web-resurrect/mcp"],
      "env": {
        "WEB_RESURRECT_API_KEY": "wr_live_xxx"
      }
    }
  }
}

Claude Code

claude mcp add web-resurrect -e WEB_RESURRECT_API_KEY=wr_live_xxx -- npx -y @web-resurrect/mcp

Available tools

The server exposes around 30 tools covering projects, pages, scraping, SEO enrichment, rewriting, images, categorization, WordPress and jobs. Examples:

• create_project, list_projects, delete_project
• scrape_page, scrape_bulk
• enrich_project (Haloscan auto + free, Majestic optional at 10 credits)
• rewrite_page, rewrite_bulk (add wisewand=true for premium quality; Wisewand also accepts Haloscan-origin pages with no snapshot)
• generate_image, generate_image_bulk
• wordpress_configure, wordpress_publish, wordpress_publish_bulk
• get_job, list_jobs, cancel_job

3. CLI

The wr CLI lets you drive the full API from your terminal. Perfect for scripts, cron jobs and automation without an AI agent.

Installation

npm install -g @web-resurrect/cli

Authentication

wr login
# or via environment variable
export WR_API_KEY=wr_live_xxx

Full workflow

# Check your balance
wr credits

# Create a project for an expired domain
wr projects create example.com --name "My project"

# Wait for URL fetching to finish
wr jobs wait <job_id>

# Enrich with Haloscan
wr enrich <project_id> --sources haloscan

# List pages sorted by traffic
wr pages list <project_id> --sort total_traffic

# Scrape pending pages
wr scrape-bulk <project_id>

# Rewrite (add --wisewand for premium)
# Also covers Haloscan-origin pages (no Wayback snapshot)
wr rewrite-bulk <project_id> --wisewand

# Generate images
wr image-bulk <project_id>

# Configure and publish to WordPress
wr wp configure https://mysite.com --mode plugin
wr wp publish <page_id> --domain mysite.com --status draft

4. Haloscan-origin pages (no Wayback snapshot)

Some URLs surface through Haloscan but have no archive on the Wayback Machine. Web Resurrect still inserts them into your project with source="haloscan". They cannot be scraped — instead they are rewritten directly with Wisewand, using the slug plus the ranked keywords from Haloscan (synthetic content is built server-side).

The source field

Every page returned by the API exposes a source field:

• wayback: discovered via the Wayback Machine, scrapable as usual.
• haloscan: discovered via Haloscan, no snapshot — Wisewand rewrite only.

Dedicated status values + filter

• status=haloscan: Haloscan-origin pages not yet rewritten.
• status=rewritable_wisewand: union of scraped + Haloscan-origin pages (anything eligible for Wisewand).
• source=haloscan: strict filter by origin.

Rewriting Haloscan pages from the CLI

# All Wisewand-eligible pages (scraped + Haloscan)
wr rewrite-bulk <project_id> --wisewand

# Only Haloscan-origin pages (skip scraped ones)
wr rewrite-bulk <project_id> --wisewand --haloscan-only

# List a project's Haloscan pages
wr pages list <project_id> --status haloscan

Rewriting Haloscan pages via MCP

The rewrite_bulk tool with wisewand=true automatically accepts Haloscan pages. To fetch their IDs:

# Ask your agent:
list_pages project_id="..." status="rewritable_wisewand"
# then
rewrite_bulk page_ids=[...] wisewand=true

5. Credit costs

6. API Reference

For the full REST API reference (endpoints, schemas, examples), see the interactive documentation.