investigating-error-issue

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect prompt injection surface identified. The skill ingests untrusted data from error stack traces, log entries, and event properties through tools like posthog:execute-sql and posthog:query-logs (found in Step 2, 3, 4, and 5 of SKILL.md). This data is subsequently synthesized into a summary in Step 6. 1. Ingestion points: Event data and logs entered into the agent context via PostHog query tools. 2. Boundary markers: Absent; there are no explicit delimiters or instructions to the agent to treat the retrieved data as untrusted. 3. Capability inventory: The skill's actions are restricted to PostHog's analytics environment, with no access to the local filesystem or external network. 4. Sanitization: No validation or sanitization of the external data is performed before synthesis.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 01:54 PM
Security Audit — agent-trust-hub — investigating-error-issue