running-marketing
Warn
Audited by Snyk on Mar 26, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.80). The skill's required workflow (e.g., Task 6: "check feed for posts" on LinkedIn, Task 7: "identify...influencers" from LinkedIn/YouTube/Reddit, Task 8: "research the show — listen to 2-3 episodes", and Task 11: "join existing communities (Slack, Discord, Reddit)") explicitly expects the agent to read and act on public, user-generated third‑party content, which could carry indirect prompt-injection instructions.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata