counter-narrative

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

• Third-party content exposure detected (high risk: 1.00). The skill's SKILL.md requires users to supply "Evidence" as URLs or screenshots of competitor content (e.g., social posts, landing pages, press releases) and the Process explicitly instructs the agent to "examine the evidence provided" to drive analysis and strategy, thereby ingesting untrusted public third‑party content that can influence the agent's decisions.