appsignal

Pass

Audited by Gen Agent Trust Hub on Apr 29, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the Membrane CLI package (@membranehq/cli@latest) from the official npm registry. This is a standard installation procedure for the vendor's own tooling.
  • [COMMAND_EXECUTION]: The instructions involve running several membrane CLI commands for authentication, connection management, and executing actions. These are legitimate operations intended for interacting with the Membrane platform.
  • [CREDENTIALS_UNSAFE]: The skill explicitly advises against asking users for API keys or tokens, instead using a server-side connection model where Membrane manages the authentication lifecycle. This is a recommended security practice.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes data from AppSignal (incidents, errors, performance issues) via Membrane actions. While this introduces a surface for indirect prompt injection from external monitor data, it is a standard risk for integration skills and the skill does not grant excessive autonomous capabilities that would escalate the risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 29, 2026, 10:09 PM