Skills
skills/cklxx/elephant.ai/auto-skill-creation/Gen Agent Trust Hub

auto-skill-creation

Warn

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill utilizes importlib.util to dynamically load and execute generated Python modules in its testing framework. This allows for the execution of code created at runtime.
  • [COMMAND_EXECUTION]: The create function builds Python scripts and Markdown files by directly interpolating user-provided strings into f-strings. This enables potential injection of malicious logic into the generated run.py.
  • [COMMAND_EXECUTION]: The skill performs directory creation and file writing using unsanitized user input for file paths. This could lead to path traversal or accidental overwriting of files outside the intended directory.
  • [PROMPT_INJECTION]: Surface for indirect prompt injection via the create function's name and description arguments.
  • Ingestion points: args in create function (run.py).
  • Boundary markers: Absent.
  • Capability inventory: File system writes and dynamic loading (run.py, tests/test_auto_skill.py).
  • Sanitization: Absent.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 11, 2026, 02:14 AM