harness

Pass

Audited by Gen Agent Trust Hub on Apr 29, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the Membrane CLI via the official NPM registry (@membranehq/cli). This is a standard procedure for this vendor's tools.
  • [COMMAND_EXECUTION]: The skill utilizes shell commands to manage authentication (membrane login), create connections (membrane connect), and execute integration actions. These commands are scoped to the vendor's management platform.
  • [INDIRECT_PROMPT_INJECTION_SURFACE]: The skill ingests data from external Harness environments through its actions. This creates a potential surface for indirect prompt injection if malicious data is present in the Harness environment being processed, though no specific exploits are present in the skill code itself.
  • [DYNAMIC_EXECUTION]: The skill includes functionality to dynamically create and build new actions (membrane action create). This is a core feature of the Membrane platform and is used to extend the integration's capabilities on the fly.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 29, 2026, 06:12 PM
Security Audit — agent-trust-hub — harness