jaegertracing

Pass

Audited by Gen Agent Trust Hub on Apr 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches the @membranehq/cli package from the NPM registry. This is the official tool provided by the vendor for managing integrations.
  • [COMMAND_EXECUTION]: Executes shell commands to interact with the Membrane platform, enabling functions such as login, connection management, and action execution.
  • [PROMPT_INJECTION]: Identifies a surface for indirect prompt injection through the processing of trace data from external microservices.
  • Ingestion points: Data retrieved from JaegerTracing via membrane action run and membrane action list commands in SKILL.md.
  • Boundary markers: No explicit delimiters or instructions are provided to the agent to distinguish between its instructions and the data ingested from traces.
  • Capability inventory: The skill enables the agent to create and execute actions (membrane action run, membrane action create) and manage connections.
  • Sanitization: No data validation or sanitization is performed on the output of the CLI before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 28, 2026, 11:50 PM
Security Audit — agent-trust-hub — jaegertracing