skills/host0-ai/skills/host0/Gen Agent Trust Hub

host0

Pass

Audited by Gen Agent Trust Hub on Oct 18, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill enables the creation of applications that ingest and display user-controlled data and API records.
  • Ingestion points: File contents within directories published via publish.sh and record data fetched from /.h0/data/.
  • Boundary markers: Not explicitly defined in the deployment script; however, the documentation provides specific guidance on escaping record values.
  • Capability inventory: The deployment script performs file system walks, hashing, and network uploads to the host0.ai API.
  • Sanitization: The provided JavaScript documentation includes an esc function to sanitize record values before rendering them to the DOM to prevent XSS.
  • [COMMAND_EXECUTION]: The publish.sh script uses standard system utilities such as curl, jq, find, and shasum to manage the deployment lifecycle. These commands are used for their intended purposes of file processing and API interaction.
  • [EXTERNAL_DOWNLOADS]: The skill references the download of its helper script from host0.ai/publish.sh and installation via npx from the vendor's repository. These operations are associated with the official vendor infrastructure.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 18, 2026, 12:45 PM