langfuse

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes code examples that pass secret_key/public_key inline (e.g., secret_key="sk-...") and shows placing API keys directly in generated code, which encourages the LLM to handle/embed secrets verbatim.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches prompt definitions at runtime from Langfuse (host="https://cloud.langfuse.com"), and those remote prompts are compiled and injected into LLM messages so the external content directly controls agent instructions and is a required dependency.