skill-creator

Fail

Audited by Gen Agent Trust Hub on Feb 25, 2026

Risk Level: CRITICALCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill directs the agent to use local utility scripts, specifically scripts/init_skill.py and scripts/package_skill.py, for initializing and packaging skills. It also uses standard shell commands like ls to interact with the skill directory structure.\n- [PROMPT_INJECTION]: The skill functions as a generator for other instructions, creating an indirect prompt injection surface by ingesting and processing untrusted user input.\n
  • Ingestion points: User input is collected for skill metadata (name, description) and instruction body content (SKILL.md, Step 4).\n
  • Boundary markers: No explicit delimiters or isolation instructions are defined to separate user-provided content from the agent's instructions in the generated output.\n
  • Capability inventory: The workflow includes writing to the local file system and executing local management scripts.\n
  • Sanitization: The instructions focus on structural organization and token efficiency but do not prescribe specific validation or sanitization protocols for the user-provided skill content.
Recommendations
  • Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
CRITICAL
Analyzed
Feb 25, 2026, 09:04 AM