signals-scout-ai-observability
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted external data in the form of LLM traces, creating a potential vector for indirect prompt injection.
- Ingestion points: The
query-llm-tracetool retrieves full request and response message content from theposthog.ai_eventstable, andexecute-sqlpulls raw event properties. - Boundary markers: There are no explicit instructions or delimiters provided to the agent to distinguish between trace data and system instructions.
- Capability inventory: The skill utilizes
scout-emit-report,scout-edit-report, andscout-scratchpad-rememberto generate reports and persist information within the PostHog project environment. - Sanitization: The instructions do not include steps for sanitizing or validating the content of the traces being analyzed, allowing potentially malicious embedded instructions to enter the agent's context.
Audit Metadata