abtesting-web

Pass

Audited by Gen Agent Trust Hub on Mar 7, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: Uses CLI tools openclaw and wrangler for managing A/B test configurations and deploying Cloudflare Workers. These are expected administrative operations for this skill's functionality.
  • [EXTERNAL_DOWNLOADS]: References official API endpoints for LaunchDarkly and Unleash for feature flag management and requires the launchdarkly-node-server-sdk library. These originate from well-known and trusted services.
  • [PROMPT_INJECTION]: Processes untrusted user data from request headers and cookies within Cloudflare Workers to implement routing logic. While this establishes a surface for indirect prompt injection, it is a standard implementation pattern for traffic segmentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 7, 2026, 05:44 PM