aiconfig-online-evals

Pass

Audited by Gen Agent Trust Hub on Apr 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill facilitates communication with the official LaunchDarkly API endpoints at https://app.launchdarkly.com/api/v2. All network operations described are consistent with the intended functionality provided by the vendor.
  • [SAFE]: Instructions for API key detection are limited to standard practices, such as checking environment variables and the local MCP configuration file (~/.claude/config.json). These are legitimate locations for service-specific credentials provided by the user to the agent environment.
  • [SAFE]: External code references and documentation links point exclusively to official, vendor-owned GitHub repositories (github.com/launchdarkly/*), which are trusted sources for integration examples.
  • [SAFE]: The skill uses standard, well-maintained libraries (requests, ldclient, ldai) for its Python implementation examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 29, 2026, 05:57 PM