two-agent-harness

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

• Third-party content exposure detected (high risk: 0.80). The skill includes explicit migration and sync code that fetches issues from external JIRA instances and similar project-management sources (see migrate_jira_features / sync_to_jira in references/feature-tracking-system and migrate scripts), which would ingest untrusted, user-generated content that the agents are expected to read and interpret.