finding-replay-for-issue
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from error logs and session metadata, which creates a potential surface for indirect prompt injection.
- Ingestion points: External data enters the context through
posthog:query-error-tracking-issue(issue names and descriptions) andposthog:execute-sql(user-controlled URLs and exception properties). - Boundary markers: The instructions do not define explicit delimiters to separate this external data from the agent's instructions.
- Capability inventory: The skill can execute SQL queries and trigger AI-based session summarization scanners (
posthog:vision-scanners-scan-session). - Sanitization: There is no explicit sanitization logic defined for the external strings before they are presented to the user or used in ranking logic.
Audit Metadata