langsmith-observability

Warn

Audited by Snyk on Feb 16, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (high risk: 0.80). The skill includes calls to pull prompts from a shared prompt hub (e.g., client.pull_prompt("langchain-ai/rag-prompt") and client.pull_prompt("my-org/qa-prompt")), which fetches externally authored hub prompts (potentially public or user-generated) that the agent then invokes and interprets, exposing it to untrusted third-party content and indirect prompt injection.

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

  • Potentially malicious external URL detected (high risk: 0.80). The skill demonstrates pulling prompts at runtime via the LangSmith hub (client.pull_prompt) which causes the agent to execute externally-fetched prompt instructions from the LangSmith API endpoint (e.g. https://api.smith.langchain.com), so remote content directly controls agent behavior.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 16, 2026, 12:58 AM