rasa
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the
@membranehq/clipackage via npm and utilizesnpxfor executing tasks. These components are official tools provided by the skill's vendor ecosystem.\n- [COMMAND_EXECUTION]: The skill's functionality is centered on executingmembraneshell commands to manage user sessions, connect to external services, and run API actions.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external Rasa assistants, presenting a surface for indirect prompt injection.\n - Ingestion points: Data retrieved from Rasa assistant trackers and events through
membrane action runormembrane request.\n - Boundary markers: The instructions do not define explicit delimiters or instructions to ignore embedded commands within the retrieved Rasa data.\n
- Capability inventory: The agent has access to shell execution via the
membraneCLI and can perform proxied network requests.\n - Sanitization: No specific sanitization or validation logic for the external API content is defined in the skill documentation.
Audit Metadata