mckinsey-market-research-deck
Pass
Audited by Gen Agent Trust Hub on Aug 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
Bashtool to perform various operational tasks. This includes copying a rendering engine from a local skill directory, running Python scripts (python3 deck_engine.py), and using headless Google Chrome to export HTML to PDF. It also utilizes system utilities likepdftoppmand ImageMagick (magick) for quality control and image processing. - [EXTERNAL_DOWNLOADS]: The skill gathers market intelligence by using
WebSearchandWebFetchto query competitor prices, market size estimates, and customer reviews from third-party websites. - [PROMPT_INJECTION]: The skill processes untrusted external data which creates a potential surface for indirect prompt injection.
- Ingestion points: Data collected via
WebSearchandWebFetchis stored in a JSON file and subsequently processed by an AI agent for verification and deck generation. - Boundary markers: The instructions do not specify the use of clear delimiters or boundary markers to separate external content from the agent's core instructions.
- Capability inventory: The skill has the ability to write to the file system and execute shell commands through the
Bashtool. - Sanitization: The workflow includes a numerical verification step using an 'adversarial verifier' agent, but it lacks specific filtering or sanitization rules for qualitative data (such as user quotes) that could contain malicious instructions.
Audit Metadata