langsmith-observability

Pass

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [SAFE]: The skill serves as a technical guide for the LangSmith observability platform. Its content is educational and aligns with standard development practices for LLM operations.\n- [EXTERNAL_DOWNLOADS]: The skill references the langsmith Python package and the langchain-ai Hub organization, which are recognized as trusted and well-known sources in the AI development community.\n- [SAFE]: The surface for indirect prompt injection was evaluated. Ingestion occurs through evaluation datasets and traced outputs (SKILL.md). Capabilities include filesystem access via the dangerously_allow_filesystem flag (advanced-usage.md). Sanitization is addressed through instructional examples demonstrating sensitive data redaction (SKILL.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 14, 2026, 08:30 AM