diagnose

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

• Third-party content exposure detected (high risk: 0.85). The required workflow ingests outsider-authored free text when --issue <N> is used: the skill pulls the GitHub issue body via gh issue view <N> and writes it into the session file (.claude/debug/<slug>.md), which the debugger subagent then reads as LLM context.