cold-email-outreach

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external lead data (e.g., names, company names) from a Supabase database to generate email content. This creates a potential surface for indirect prompt injection where malicious data in the source database could attempt to influence the agent's behavior.
  • Ingestion points: Data from the people table in Supabase accessed in Phase 1.
  • Boundary markers: The instructions do not define specific delimiters to separate lead data from system instructions during email generation.
  • Capability inventory: The skill possesses file-writing capabilities (CSV export), database-writing capabilities (logging to Supabase), and network access via MCP tools for Smartlead.
  • Sanitization: No explicit sanitization or filtering of lead-sourced strings is mentioned.
  • [COMMAND_EXECUTION]: Documentation includes a setup command (python3 tools/supabase/setup_database.py) intended for initializing the database schema. This is a routine administrative task for the skill's operation.
  • [EXTERNAL_DOWNLOADS]: The skill references and interacts with several well-known third-party services and tools (Supabase, Smartlead, Instantly, Lemlist, Apollo) which are necessary for its documented cold email outreach functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 03:05 PM
Security Audit — agent-trust-hub — cold-email-outreach