landing-page-design
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions to execute AI models (such as image generators and search assistants) using the
beltCLI tool, which is limited by the skill's allowed tools scope. - [EXTERNAL_DOWNLOADS]: The skill references an external installation script for the
beltCLI located athttps://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md. - [INDIRECT_PROMPT_INJECTION]: The skill integrates with external search tools to gather competitive intelligence and statistics, creating a surface for indirect prompt injection from untrusted web content.
- Ingestion points: External data enters the agent context via the output of
tavily/search-assistantandexa/answertools (SKILL.md). - Boundary markers: There are no explicit delimiters or instructions to ignore instructions embedded in the search results.
- Capability inventory: The skill is configured to run arbitrary applications within the
beltCLI ecosystem (SKILL.md). - Sanitization: No sanitization or validation of the retrieved search content is described in the skill instructions.
Audit Metadata