onboarding

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill explicitly installs and loads companion skills from the public GitHub repo (npx skills add launchdarkly/ai-tooling) and repeatedly instructs the agent to call MCP/ldcli/API tools (e.g., list-feature-flags, get-environments, create-feature-flag) that fetch live, user/third-party LaunchDarkly data which the agent is expected to read and act on as part of its workflow, so untrusted external content can materially influence decisions and tool use.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The onboarding flow explicitly runs "npx skills add launchdarkly/ai-tooling" during Steps 0-3, pulling code from https://github.com/launchdarkly/ai-tooling at runtime which installs skill definitions that directly control agent prompts/behavior and executes remote code, so this is a runtime external dependency I flag.