landing-page-design

Fail

Audited by Gen Agent Trust Hub on Aug 26, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill provides installation instructions that use a highly risky pattern: curl -fsSL https://cli.inference.sh | sh. This command downloads a script from a remote server and pipes it directly into the shell for execution, bypassing local verification and potentially allowing for arbitrary code execution if the source is compromised.
  • [INDIRECT_PROMPT_INJECTION]: The skill includes instructions to use search tools that ingest untrusted data from the web, creating a vulnerability to indirect prompt injection.
  • Ingestion points: External data is ingested from the web via search tools in SKILL.md (e.g., tavily/search-assistant and exa/answer).
  • Boundary markers: The skill does not employ any boundary markers, delimiters, or explicit instructions to ignore embedded commands within the search results.
  • Capability inventory: The agent has access to the infsh CLI which can perform network operations, execute search queries, and generate images.
  • Sanitization: There is no evidence of sanitization, filtering, or validation of the data returned from the search APIs before it is processed by the agent.
  • [EXTERNAL_DOWNLOADS]: The skill attempts to install external dependencies using npx skills add inference-sh/skills@.... These packages are hosted on external registries and represent unverified third-party dependencies.
Recommendations
  • HIGH: Downloads and executes remote code from: https://cli.inference.sh - DO NOT USE without thorough review
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Aug 26, 2026, 11:04 PM
Security Audit — agent-trust-hub — landing-page-design