ib-check-deck
Pass
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes text content from external PowerPoint presentations, which serves as an ingestion point for potentially untrusted data. This introduces a surface for indirect prompt injection, where malicious instructions embedded in a deck could attempt to influence the agent's behavior.\n
- Ingestion points: Slide content from live presentations or uploaded
.pptxfiles as specified inSKILL.md.\n - Boundary markers: The instructions do not define delimiters or specific constraints to prevent the agent from following instructions found within the deck text.\n
- Capability inventory: The skill involves writing to a temporary file (
/tmp/deck_content.md) and executing a local script (scripts/extract_numbers.py).\n - Sanitization: There is no evidence of text sanitization or filtering applied to the slide content before it is processed by the agent or passed to the analysis script.\n- [COMMAND_EXECUTION]: The skill executes a local Python script (
scripts/extract_numbers.py) to perform numerical analysis on the presentation data. While this is a intended functional capability, it involves interaction with the host's command-line interface.
Audit Metadata