CRED-1: An Open Multi-Signal Domain Credibility Dataset for Automated Pre-Bunking of Online Misinformation

What's New

🤖 MCP Server (Model Context Protocol)

New cred1-mcp binary — 5 tools for AI agents to check source credibility:

| Tool | Description | |------|-------------| | check_domain | Check a single domain's credibility score | | batch_check | Check up to 100 domains at once | | search_domains | Search domains by pattern/substring | | get_stats | Dataset statistics (2,673 domains, 6 categories) | | get_categories | Category taxonomy with descriptions and scores |

Configure in Claude Desktop / Cursor / Windsurf

{
  "mcpServers": {
    "cred1": {
      "command": "npx",
      "args": ["-y", "@aloth/cred1", "--mcp"]
    }
  }
}

📦 npm Package

• Library + CLI: npm install -g @aloth/cred1
• Quick check: npx @aloth/cred1 check infowars.com
• Programmatic: import { checkDomain } from '@aloth/cred1'

Dataset

• 2,673 domains scored 0.0–1.0
• Categories: fake, conspiracy, unreliable, satire, mixed, reliable
• Weekly updates via GitHub Actions