langchain-webhooks-events

Pass

Audited by Gen Agent Trust Hub on Mar 24, 2026

Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill defines a 'WebhookCallbackHandler' using the 'httpx' library to send event data, including prompts and model outputs, to a user-specified webhook URL. This is consistent with the skill's stated purpose of implementing event-driven LangChain applications.- [PROMPT_INJECTION]: The provided FastAPI code snippets for WebSocket and SSE endpoints establish a surface for indirect prompt injection by processing external input.
  • Ingestion points: The 'message' parameter in 'stream_chat' and 'data["message"]' in 'websocket_chat'.
  • Boundary markers: The code lacks delimiters or explicit instructions to the LLM to ignore embedded commands in the user input.
  • Capability inventory: The skill enables external network requests via webhooks and LLM invocation.
  • Sanitization: No input sanitization or validation logic is provided in the implementation examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 24, 2026, 04:31 PM