skill-installer
Warn
Audited by Snyk on Feb 15, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.90). This skill fetches and installs skills from public GitHub (e.g., the curated list at https://github.com/openai/skills/... and arbitrary user-provided GitHub repo/URLs), so it ingests untrusted third‑party code and listings that the agent will read/interpret and install, enabling indirect prompt injection.
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.90). This skill installer fetches and installs executable skill code at runtime from GitHub (e.g. https://github.com///tree// and the curated list at https://github.com/openai/skills/tree/main/skills/.curated), which will place remote code into $CODEX_HOME/skills that can directly control agent behavior or execute code.
Audit Metadata