skill-creator

Pass

Audited by Gen Agent Trust Hub on Oct 8, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely instructional and does not contain executable code, dangerous system commands, or persistence mechanisms.
  • [SAFE]: The instructions explicitly promote high-security standards by warning agents not to include unrestricted network tools like 'Bash(curl:*)' and to always place security warnings directly next to high-risk configurations.
  • [SAFE]: All external URLs link to official 'apollographql.com' documentation or the 'agentskills.io' specification, with no evidence of typosquatting or homoglyph attacks.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a functional surface for processing user-supplied skill descriptions and requirements. However, it mitigates potential injection risks by instructing the agent to never guess security settings and to always require explicit user confirmation for data-sensitive features.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 8, 2026, 10:36 PM