persona-evaluator

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

• Third-party content exposure detected (high risk: 0.90). The skill's SKILL.md and references/BLACK-BOX.md explicitly instruct the agent to ingest and rubric-score third-party content—including A2A "pack-content" responses and Tier 3 "passive observation" of public broadcasts/Slack/forums—which the agent must read and interpret to produce evaluation reports, so untrusted user-generated content can materially influence its decisions and actions.