skills/canhta/deliverable/upgrade/Gen Agent Trust Hub

upgrade

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches versioning information and source code from the author's GitHub repository (github.com/canhta/deliverable) to facilitate the update process.
  • [COMMAND_EXECUTION]: Utilizes system commands such as git, curl, and npx to manage installation detection and file synchronization. These operations are limited to the skill's defined installation directories.
  • [PROMPT_INJECTION]: Ingestion point: Remote version data from raw.githubusercontent.com (SKILL.md); Boundary markers: Absent; Capability inventory: Shell execution via git, npx, and curl (SKILL.md); Sanitization: Absent. The surface for indirect injection exists but the source is the author's own repository.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 07:17 AM