campaign-sending

Warn

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to fetch provider API documentation from external websites if a pre-configured reference is not available. This retrieval of unverified content from the internet introduces risk.
  • [DATA_EXFILTRATION]: The skill's primary function is to transmit sensitive data, including API keys and contact information (emails, names, company domains), to external service providers such as Instantly.ai.
  • [COMMAND_EXECUTION]: The workflow involves the execution of network requests and API interactions to manage campaigns and upload leads to remote servers.
  • [PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection:
  • Ingestion points: Processes external CSV files containing lead data and potentially untrusted API documentation fetched from the web.
  • Boundary markers: The instructions do not specify the use of delimiters or 'ignore' commands when processing the external data.
  • Capability inventory: The agent performs network operations (API calls) which could be influenced by malicious instructions embedded in the ingested content.
  • Sanitization: No explicit sanitization or validation steps are defined for the documentation content or CSV fields before they are used to drive the workflow.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 3, 2026, 09:41 PM