create-command

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface.
  • Ingestion points: User-provided <command-name> and <description> are used to construct the content of new markdown files in the local filesystem.
  • Boundary markers: The skill does not implement delimiters or instructions for the agent to ignore content within the generated files.
  • Capability inventory: The skill utilizes file-writing capabilities to create files in .cursor/commands/ and .claude/commands/trellis/.
  • Sanitization: No validation or escaping is applied to the user-supplied description before it is written to the markdown files, which could allow a user to inject malicious instructions that are interpreted by the agent during future command execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 10:15 AM