exploring-ai-failures
Warn
Audited by Snyk on Jul 7, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.85). SKILL.md instructs the workflow to “read real traces” via
posthog:query-llm-trace/posthog:query-llm-traces-list, which at runtime will ingest trace content authored by external end-users (e.g., their prompts/messages) into the agent’s LLM context for analysis.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata