content-strategy

Pass

Audited by Gen Agent Trust Hub on May 13, 2026

Risk Level: SAFE
Full Analysis
  • Financial Data Access: The skill is designed to pull transaction history and revenue data from QuickBooks, PayPal, and Square. While this involves handling sensitive business information, the access is performed through authenticated connectors for the explicit purpose of generating marketing insights.
  • Potential for Indirect Prompt Injection: By processing external transaction data (such as product names or descriptions), the skill possesses an attack surface for indirect prompt injection. This is a potential consideration when an agent ingests data from external sources. However, as the skill's output is restricted to strategic recommendations and does not execute system-level commands, the associated risk is minimal.
  • Ingestion points: SKILL.md (Step 3: Pull and analyze sales data) describes fetching data from QuickBooks, PayPal, and Square.
  • Boundary markers: No specific delimiters or "ignore instructions" markers are defined for the data analysis phase.
  • Capability inventory: The skill uses financial data fetching tools (profit-loss-quickbooks-account, list_transactions) and returns structured JSON output. No dangerous capabilities like shell access or file writes were identified.
  • Sanitization: No explicit sanitization of transaction data is documented.
  • Tool Usage and Intent: The identified tool calls, such as list_transactions and make_api_request, align with the skill's stated purpose of financial analysis. The workflow demonstrates a clear path from data retrieval to human-in-the-loop approval, which is a security best practice.
Audit Metadata
Risk Level
SAFE
Analyzed
May 13, 2026, 05:33 PM