adr

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

• Potentially malicious external URL detected (high risk: 0.90). The skill fetches ADR page bodies at runtime from the Confluence REST API (e.g., https://{ATLASSIAN_SITE}/wiki/rest/api/content/... and https://{ATLASSIAN_SITE}/wiki/api/v2/pages/...) and injects that fetched content directly into the sub-agent review prompt, so remote content can control agent instructions.