create-skill

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest user input through a conversational interview and write those inputs into new instruction files (SKILL.md).
  • Ingestion points: User responses gathered during the three-question interview described in Step 1 of SKILL.md.
  • Boundary markers: No specific delimiters are mandated for user input, though the skill provides structured sections (Steps, Gotchas, etc.) to house the information.
  • Capability inventory: The skill is intended to perform file system write operations to ~/.claude/skills/ or .claude/skills/ as specified in Step 2 and Step 6 of SKILL.md.
  • Sanitization: The instructions focus on content quality and triggering accuracy rather than input sanitization, which is expected for a developer-oriented creation tool.
  • Assessment: This represents the core functional surface of a meta-skill and does not pose a threat in this context.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 11:20 PM
Security Audit — agent-trust-hub — create-skill