investigating-logs
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of instructional content for using PostHog's MCP tools to analyze logs. It encourages best practices such as summarizing data before inspecting raw rows and scoping queries to specific services. No executable code, remote downloads, or malicious instructions were found.
- [INDIRECT_PROMPT_INJECTION]: The skill processes log data, which acts as an untrusted external data source. 1. Ingestion points: posthog:query-logs and posthog:logs-patterns tools. 2. Boundary markers: None mentioned for log data. 3. Capability inventory: PostHog data analysis tools and SQL execution. 4. Sanitization: No explicit sanitization of log contents mentioned.
Audit Metadata