engineering-advanced-skills

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

• Third-party content exposure detected (medium risk: 0.30). In the agenthub/skills/eval runtime workflow’s LLM-judge mode, the coordinator reads each agent’s result post from .agenthub/board/results/agent-{i}-result.md and diffs from git diff base...agent_branch, so outsider-authored content can enter via the agent-generated board files.