skills/langwatch/skills/analytics/Gen Agent Trust Hub

analytics

Pass

Audited by Gen Agent Trust Hub on Apr 25, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill exclusively utilizes Model Context Protocol (MCP) tools specifically designed for querying analytics and observing agent performance.\n- [SAFE]: The instructions contain strong negative constraints, explicitly forbidding the agent from writing code, installing SDKs, or using tools to create resources, which limits the operational scope to data retrieval.\n- [SAFE]: No obfuscated content, hardcoded credentials, or suspicious network activities were detected in the skill markdown.\n- [SAFE]: The skill interacts with external data by retrieving individual traces for inspection. While these traces contain untrusted user inputs, the read-only nature of the available tools prevents malicious exploitation.\n
  • Ingestion points: External trace details and request/response logs retrieved through the get_trace and search_traces tools.\n
  • Boundary markers: No specific delimiters are used to wrap the retrieved trace data.\n
  • Capability inventory: Access is limited to read-only metrics and trace retrieval tools (get_analytics, search_traces, get_trace).\n
  • Sanitization: The skill does not define specific sanitization routines, relying on the agent's summarization instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 25, 2026, 06:42 PM
Security Audit — agent-trust-hub — analytics