@web-resurrect/mcp

MCP (Model Context Protocol) server for the Web Resurrect API. Lets AI assistants like Claude resurrect expired domains: fetch archived URLs, scrape content, rewrite with AI, generate images, and publish to WordPress.

🤖 AI Agent Skill (pipeline complet)

Le skill unifié `web-resurrect` couvre tout le workflow de A à Z (création projet → enrichissement SEO → scraping Wayback → réécriture → catégorisation AI → publication WordPress → redirections), avec toutes les astuces et pièges connus. Il fonctionne aussi bien avec ce MCP server qu'avec la CLI wr, et contient une table de correspondance complète entre les deux modes.

Compatible avec Claude Code, Codex, Cursor, Cline, Copilot, OpenCode, Windsurf et 40+ autres agents via npx skills :

# Install globalement pour tous tes projets
npx skills add MattiooFR/web-resurrect-skill -g

# Ou uniquement pour le projet courant
npx skills add MattiooFR/web-resurrect-skill

Une fois installé, demande simplement à ton agent :

"Ressuscite le domaine exemple.com"

Et il suivra automatiquement le pipeline complet en utilisant les outils MCP mcp__web_resurrect__* de ce serveur.

Source du skill : github.com/MattiooFR/web-resurrect-skill

Quick start

Claude Desktop

Add 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

Local development

cd packages/mcp
npm install
npm run build
WEB_RESURRECT_API_KEY=wr_live_xxx node dist/index.js

Environment variables

Available tools

Credits

• get_credits — Get current credit balance and account info

Projects

• create_project — Create a project for an expired domain (starts URL fetching)
• list_projects — List all projects with page counts
• get_project — Get project details and stats
• delete_project — Delete a project and all its pages

Pages

• list_pages — List pages with status/source/search filters, sorting, pagination. Each page exposes a source field: wayback (default, scrapeable from archive) or haloscan (discovered via Haloscan, no snapshot — Wisewand-rewrite only).
• get_page — Get full page details (scrape, rewrite, SEO, WordPress status)

Scraping

• scrape_page — Scrape a page from the Wayback Machine (1 credit)
• scrape_bulk — Scrape multiple pages (1 credit each, max 100)

SEO Enrichment

• enrich_project — Enrich with Haloscan (free) and/or Majestic (10 credits)

Rewriting

• rewrite_page — Rewrite a page (basic default 1 credit, add wisewand=true for premium 5 credits / 1 with own key). Wisewand mode also accepts Haloscan-origin pages (no scrape required).
• rewrite_bulk — Rewrite multiple pages (same options, max 50). In Wisewand mode, mixes scraped + Haloscan-origin pages; use list_pages with status='rewritable_wisewand' to fetch everything rewritable in one call, or status='haloscan' to target only Haloscan-origin pages.

Image Generation

• generate_image — Generate AI featured image (1 credit)
• generate_image_bulk — Generate images for multiple pages (1 credit each, max 50)

Categorization

• categorize_page — AI-suggest a WordPress category for a page (free)

WordPress

• wordpress_plugin_check — Check if WP plugin is installed
• wordpress_configure — Configure WordPress credentials (plugin or basic_auth)
• wordpress_validate — Validate WordPress connection
• wordpress_categories — List WordPress categories
• wordpress_authors — List WordPress authors
• wordpress_publish — Publish a page to WordPress (free, original URLs preserved in plugin mode)
• wordpress_publish_bulk — Publish multiple pages (free, supports plugin and Basic Auth modes)
• export_redirects — Export URL mappings to file (only needed for Basic Auth mode)

Jobs

• get_job — Get job status, progress, and result
• wait_for_job — Block until a job reaches a terminal state (server-side polling). USE THIS AFTER EVERY ASYNC CALL for autonomous pipelines — no more manual polling loops.
• list_jobs — List recent jobs with filters
• cancel_job — Cancel a pending job

Overview

• get_project_overview — Single-call pipeline status: pending/scraped/rewritten/published counts, Wisewand pipeline state, SEO totals. Replaces several list_pages calls.

Typical workflow

1. create_project with a domain -> get job_id
2. get_job to wait for URL fetching to complete
3. enrich_project with Haloscan for SEO data
4. list_pages sorted by total_traffic to find best pages
5. scrape_bulk the top pages
6. rewrite_bulk the scraped pages
7. generate_image_bulk for rewritten pages
8. wordpress_configure your WordPress site
9. wordpress_publish_bulk to publish everything (original URLs preserved in plugin mode, URL mappings pushed automatically)
10. Non-published URLs automatically 301 to homepage (via plugin catch-all)

Credit costs