exec-summary

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes data from external sources including Google Analytics, Meta Ads, HubSpot, and Salesforce.\n
  • Ingestion points: Data enters through connected MCPs and local profile files in ~/.claude-marketing/.\n
  • Boundary markers: The instructions lack explicit delimiters or safety prompts to ignore commands embedded in the marketing data.\n
  • Capability inventory: The skill can execute local scripts and dispatch formatted messages via Slack or email.\n
  • Sanitization: There is no evidence of input validation or sanitization before processing the untrusted data.\n- [COMMAND_EXECUTION]: The skill executes local Python scripts (clv-calculator.py, roi-calculator.py, revenue-forecaster.py, and budget-optimizer.py) to calculate KPIs and generate forecasts. These scripts are invoked as part of the core analytical process.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 12:41 AM
Security Audit — agent-trust-hub — exec-summary