Skills
skills/myuon/agent-skills/create-agent-skills/Gen Agent Trust Hub

create-agent-skills

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection because it instructs the agent to create new skill files based on $ARGUMENTS without providing guidance on sanitization or boundary markers.
  • Ingestion points: User-supplied input via $ARGUMENTS in SKILL.md.
  • Boundary markers: None mentioned; the guide does not instruct the agent to wrap generated content in delimiters or include warnings to ignore embedded instructions.
  • Capability inventory: The skill is granted Write, Edit, and Bash tools, which are used to commit the generated content to the filesystem.
  • Sanitization: Absent; the guide lacks instructions for escaping or validating the content provided in $ARGUMENTS before writing it to a new SKILL.md file.
  • [EXTERNAL_DOWNLOADS]: The documentation references and encourages the use of an external tool, skills-ref, for skill validation. This tool is hosted on a third-party GitHub repository (github.com/agentskills/agentskills).
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 08:51 AM