Veritier Fact Checking

Real-time fact-checking and claim extraction. Extract falsifiable claims from any text or document, then verify each against live web evidence or your own pr...

This appears to be a coherent Veritier fact-checking integration. Before installing or running examples: 1) Only provide a vt_ API key to the veritier.ai endpoint and keep test keys (vt_test_...) separate from production keys. 2) If you run the Python/JS examples, install dependencies in an isolated environment (venv/container) and review the code yourself; the examples make outbound calls to api.veritier.ai. 3) Be cautious with logs and CI — test_integration.py prints the first 12 characters of the key (local console output), so avoid running it in shared logs with production keys. 4) If you plan to accept webhooks, protect your VERITIER_WEBHOOK_SECRET and deploy the receiver over HTTPS as suggested. 5) Only install the skill if you trust veritier.ai for handling your verification requests; you can revoke API keys from the dashboard at any time.

Type: OpenClaw Skill Name: veritier-fact-checking Version: 2.1.2 The Veritier fact-checking skill bundle is a well-documented and legitimate integration for an AI-powered verification service. The code, including the Python MCP proxy (veritier_mcp_proxy.py) and various JavaScript/Python examples, follows security best practices by hardcoding the API destination (https://api.veritier.ai) to prevent credential exfiltration and using timing-safe HMAC-SHA256 verification for webhooks (webhook_receiver.py). The instructions in SKILL.md are appropriately scoped to user onboarding and error handling, and the inclusion of a dedicated 'test mode' using 'vt_test_' keys demonstrates a focus on safe developer integration without malicious intent.