skills/jeremylongshore/tons-of-skills-marketplace/langfuse-reference-architecture/Gen Agent Trust Hub
langfuse-reference-architecture
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides purely educational and architectural content for Langfuse observability. It includes code snippets that demonstrate best practices such as utilizing environment variables for credentials (e.g.,
LANGFUSE_SECRET_KEY) rather than hardcoding them, and implementing graceful shutdown handlers for tracing SDKs. - [INDIRECT_PROMPT_INJECTION]: The skill provides patterns for tracing application inputs and outputs. As an observability tool, it naturally ingests external data into the trace context. While the provided examples do not explicitly include sanitization steps for the data being traced, this is a standard characteristic of logging/observability frameworks and represents a low-level surface risk common to such tools.
Audit Metadata