landing-page-design
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill directs the agent to fetch installation instructions from a remote GitHub repository (inference-sh) and to install external packages using the npx package manager (belt-sh/cli, inference-sh/skills).\n- [COMMAND_EXECUTION]: Instructs the agent to execute shell commands using the
beltCLI tool to run AI applications and search assistants.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external search tools, which provides a potential surface for indirect prompt injection attacks where malicious content in search results could influence agent behavior.\n - Ingestion points: Search results from
tavily/search-assistantandexa/answerare loaded into the agent's context.\n - Boundary markers: There are no specific delimiters or instructional barriers defined to isolate external search results from the system instructions.\n
- Capability inventory: The skill possesses command execution capabilities through the
beltCLI and general shell access provided by theBashtool.\n - Sanitization: The instructions do not specify any validation or sanitization protocols for the content retrieved from external queries.
Audit Metadata