querying-posthog-data

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

• Third-party content exposure detected (high risk: 0.80). The skill's workflow explicitly queries PostHog "events" and "document_embeddings" (e.g., references/example-llm-trace.md and references/guidelines.md) which contain user-generated/untrusted properties like $ai_input, $ai_output, and arbitrary event properties that the agent is instructed to read/inspect and which could contain embedded instructions influencing subsequent actions.