skill-creator
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the creation of new agent instructions based on user input, which is its primary purpose. This functionality represents an ingestion surface for potentially untrusted data that influences downstream behavior.\n
- Ingestion points:
SKILL.mdcapture phase for user intent and domain requirements.\n - Boundary markers: Uses structured YAML frontmatter and Markdown sections to separate instructions from metadata.\n
- Capability inventory: Instructions for the agent to use
present_filesfor skill installation, along with web search and codebase exploration tools.\n - Sanitization: The skill incorporates a mandatory 'Anti-Pattern Watchlist' to scan and refine generated instructions, providing a layer of quality control and safety steering.
Audit Metadata