bento-slides
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches the Bento application runtime (Bento_Slides.bento.html) and documentation (agents.md) from the bento.page domain.
- [COMMAND_EXECUTION]: The agent is instructed to use curl, Invoke-WebRequest, and system commands like open, start, or xdg-open to download and preview presentation files.
- [PROMPT_INJECTION]: The skill processes external user-provided content into slide data, presenting an indirect prompt injection surface.
- Ingestion points: The skill reads source material provided by the user and content from existing .bento.html files.
- Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands in the ingested data.
- Capability inventory: The agent has the ability to perform network downloads, write to the local file system, and execute files via the system's default browser handler.
- Sanitization: The instructions include a requirement to escape '<' as \u003c within the JSON block to prevent premature termination of the script tag and ensure the structural integrity of the HTML container.
Audit Metadata