deck-report-qc
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- Local Data Processing: The skill includes a dedicated Python script,
scripts/inspect_deck_report.py, which performs text and numerical extraction from various file formats such as PowerPoint, Word, and Excel. This approach is consistent with the skill's stated purpose of auditing financial deliverables and ensures that the analysis is performed locally. - Controlled Command Execution: The instructions guide the agent to execute specific local scripts to automate the QC process. These commands are well-defined and restricted to the internal skill directories, representing a standard use of automation for document analysis.
- Absence of Network Operations: The scripts and instructions were reviewed for network connectivity patterns. The analysis confirmed that no external network requests are made, which significantly reduces the risk of data exfiltration.
- Structured Output and Validation: The skill generates structured reports (HTML, CSV, JSON) and includes validation steps, such as checking for repeated metrics and source coverage. This structured approach helps ensure accuracy and prevents data manipulation during the QC process.
Audit Metadata