ai-readiness
Pass
Audited by Gen Agent Trust Hub on Jun 12, 2026
Risk Level: SAFE
Full Analysis
- Data Sensitivity Considerations: The skill is designed to analyze sensitive materials such as quarterly financials, board decks, and headcount data. These operations occur within the agent's environment, and the analysis did not identify any patterns related to unauthorized data transmission or external exfiltration.
- External Data Processing Surface: The workflow involves reading content from various file types (PDF, Excel, PPT) which can be a surface for indirect prompt injection.
- Ingestion points: Step 1 processes quarterly updates, board decks, and financial statements provided via local files or uploads.
- Boundary markers: The instructions do not explicitly define delimiters to isolate external document content from the analytical prompts.
- Capability inventory: The skill contains no executable scripts, system commands, network tools, or file-writing capabilities, which significantly limits the risk of any embedded instructions being executed.
- Sanitization: Content from the ingested materials is processed directly as text without specific sanitization or filtering logic.
- Workflow Integrity: The logic follows a clear, rule-based approach for ranking opportunities based on financial metrics and operational gates, with no signs of malicious obfuscation or behavior override attempts.
Audit Metadata