skills/hoangnb24/skills/agent-creator/Gen Agent Trust Hub

agent-creator

Pass

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a bundled Node.js script (scripts/init_agent.mjs) to generate agent configuration files. The script implements path validation using a regex for the 'slug' parameter to ensure files are only written to the '.codex/agents/' directory, preventing directory traversal.
  • [PROMPT_INJECTION]: The skill has an indirect prompt injection surface.
  • Ingestion points: Reads existing agent definitions, skill inventory, and MCP configurations from the local repository (e.g., .codex/agents/.toml, .codex/skills//SKILL.md).
  • Boundary markers: None identified to separate repository data from instructional context.
  • Capability inventory: Performs file writes to the repository and executes a local script to generate configurations.
  • Sanitization: Validates the agent 'slug' but does not sanitize the contents of the 'instructions' or 'description' fields beyond basic TOML formatting.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 18, 2026, 09:29 AM
Security Audit — agent-trust-hub — agent-creator