skills/yulonghe97/ystack/scaffold/Gen Agent Trust Hub

scaffold

Pass

Audited by Gen Agent Trust Hub on Apr 26, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the cat command to read user-specified project plan files in SKILL.md (Phase 0). This is a standard operation for a development tool to ingest local data.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from project plans to generate file content and structure.
  • Ingestion points: File reading via cat and inline text input in SKILL.md.
  • Boundary markers: Absent.
  • Capability inventory: File system read (cat) and write (.mdx, .json, .md, .ts) operations across multiple phases.
  • Sanitization: None detected; the skill relies on the underlying LLM to parse and structure freeform text input.
  • [DATA_EXPOSURE]: The skill can read any file the user provides a path to via the cat command. This is considered safe behavior in the context of an interactive developer CLI tool where the user controls the input path.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 26, 2026, 07:52 AM
Security Audit — agent-trust-hub — scaffold