fsi-strip-profile
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill requires the agent to use shell commands (
sofficeandpdftoppm) to convert generated PowerPoint presentations into images. This process is used for a mandatory visual review to check for text overlaps and layout integrity, which is a legitimate part of the skill's quality assurance workflow. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data from external sources such as SEC EDGAR filings, investor presentations, and market data providers (Bloomberg, FactSet). While this creates a potential surface for indirect prompt injection via malicious data in public filings, the risk is mitigated by the structured nature of the data collection (financial metrics, tables) and the final human-in-the-loop approval requirement before completing each slide.
- [REMOTE_CODE_EXECUTION]: The instructions include code snippets for PptxGenJS (JavaScript) and Python for slide creation. These are standard implementation examples for the agent to follow and do not involve downloading or executing code from untrusted remote servers.
Audit Metadata