technical-blog-writing

Pass

Audited by Gen Agent Trust Hub on Mar 7, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill recommends installing auxiliary components from the inference-sh organization using the npx command.
  • [COMMAND_EXECUTION]: Uses the infsh CLI to perform authentication and run various platform applications like exa/search and html-to-image.
  • [DYNAMIC_EXECUTION]: Executes Python code for chart generation via the platform's python-executor application.
  • [INDIRECT_PROMPT_INJECTION]: The skill features a search integration that processes untrusted data from the web.
  • Ingestion points: Data is ingested via the infsh app run exa/search command (SKILL.md).
  • Boundary markers: None identified in the provided implementation examples.
  • Capability inventory: The skill is authorized to use the Bash(infsh *) tool, which allows for remote application execution and network interactions (SKILL.md).
  • Sanitization: No input sanitization or filtering of search results is specified.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 7, 2026, 12:08 PM