adversus

Pass

Audited by Gen Agent Trust Hub on Mar 4, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The instructions include installing the @membranehq/cli package globally via npm, which is the official CLI tool provided by the vendor.
  • [COMMAND_EXECUTION]: The skill relies on executing various membrane shell commands to manage authentication, list actions, and run API requests.
  • [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection due to its core function of retrieving and processing external data from Adversus. 1. Ingestion points: Data is ingested from the Adversus API through actions like get-lead and raw requests using membrane request. 2. Boundary markers: The documentation does not specify the use of delimiters or 'ignore' instructions for the data retrieved from the API. 3. Capability inventory: The skill allows for data modification (write) in the Adversus platform and execution of shell commands through the CLI. 4. Sanitization: No sanitization or validation steps are defined for the content returned by the external service before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 4, 2026, 08:55 AM