langfuse-core-workflow-a

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the creation of LLM workflows (such as RAG pipelines and OpenAI wrappers) that ingest untrusted user data as variables (e.g., query and prompt parameters in Steps 1, 2, 3, 5, and 6).
  • Ingestion points: Found in code templates within SKILL.md (Steps 1, 2, 3, 5, 6).
  • Boundary markers: The provided code snippets do not include explicit delimiters or "ignore embedded instructions" warnings for the user-supplied content.
  • Capability inventory: The skill utilizes Bash(npm:*) for package management and script execution, which can execute network requests to LLM and observability providers.
  • Sanitization: The implementation templates do not demonstrate input validation or escaping for the external content interpolated into prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 03:48 AM
Security Audit — agent-trust-hub — langfuse-core-workflow-a