skill-creator

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill functions as a template generator that processes user-provided trigger words and descriptions to create new instruction files. This represents a surface for indirect prompt injection.\n
  • Ingestion points: User requests for creating or updating skills as described in SKILL.md.\n
  • Boundary markers: Enforcement of a specific YAML frontmatter and markdown section order in the generated files.\n
  • Capability inventory: Instructions to create and modify local files in the skills/ directory and update AGENTS.md.\n
  • Sanitization: The skill follows a normative style guide but does not explicitly detail sanitization of user-provided strings within the generated files.\n- [SAFE]: The skill operates entirely on local project files and follows standard development practices without introduces unauthorized command execution or data exfiltration.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 03:17 PM
Security Audit — agent-trust-hub — skill-creator